Update demo replay validation and testing documentation

- Modified the autodev state to reflect the current testing phase and details of the new `jetson-e2e` tests.
- Enhanced the "How to Test" documentation to provide clearer instructions on the demo replay validation process, including video and tlog alignment steps.
- Updated architectural documentation to include the new demo replay operator flow and its dependencies.
- Documented the removal of deprecated auto-sync features and clarified the operator-facing UI for replay validation.
- Added new entries in the dependencies table for upcoming tasks related to the demo replay flow.

These changes improve clarity and usability for operators and developers working with the demo replay system.

.clang-format

@@ -0,0 +1,18 @@
+---
+BasedOnStyle: Google
+ColumnLimit: 100
+IndentWidth: 4
+TabWidth: 4
+UseTab: Never
+AccessModifierOffset: -4
+AllowShortFunctionsOnASingleLine: Empty
+AllowShortIfStatementsOnASingleLine: false
+AllowShortLoopsOnASingleLine: false
+BinPackArguments: false
+BinPackParameters: false
+BreakBeforeBraces: Attach
+DerivePointerAlignment: false
+PointerAlignment: Left
+SortIncludes: true
+SpaceAfterCStyleCast: true
+Standard: c++17

.clang-tidy

@@ -0,0 +1,18 @@
+---
+Checks: >
+  -*,
+  bugprone-*,
+  clang-analyzer-*,
+  cppcoreguidelines-*,
+  modernize-*,
+  performance-*,
+  readability-*,
+  -bugprone-easily-swappable-parameters,
+  -cppcoreguidelines-avoid-magic-numbers,
+  -cppcoreguidelines-non-private-member-variables-in-classes,
+  -modernize-use-trailing-return-type,
+  -readability-identifier-length,
+  -readability-magic-numbers
+WarningsAsErrors: ''
+HeaderFilterRegex: '.*'
+FormatStyle: file

.cmake-format.yaml

@@ -0,0 +1,7 @@
+format:
+  line_width: 100
+  tab_size: 2
+  use_tabchars: false
+  separate_ctrl_name_with_space: false
+  separate_fn_name_with_space: false
+  dangle_parens: false

.cursor/README.md

@@ -11,10 +11,20 @@ If you want to run a specific skill directly (without the orchestrator), use the
 ```
 /problem                — interactive problem gathering → _docs/00_problem/
 /research               — solution drafts → _docs/01_solution/
-/plan                   — architecture, components, tests → _docs/02_document/
-/decompose              — atomic task specs → _docs/02_tasks/todo/
-/implement              — batched parallel implementation → _docs/03_implementation/
-/deploy                 — containerization, CI/CD, observability → _docs/04_deploy/
+/plan                   — architecture, ADRs, components, tests, epics → _docs/02_document/
+/test-spec              — blackbox/perf/resilience/security test specs → _docs/02_document/tests/
+/decompose              — atomic task specs (multi-mode) → _docs/02_tasks/todo/
+/implement              — sequential dependency-aware batches with code review and completeness gates → _docs/03_implementation/
+/test-run               — runs the test suite (functional / perf modes) with gating
+/code-review            — multi-phase review used by /implement
+/refactor               — 8-phase structured refactoring (incl. testability sub-mode) → _docs/04_refactoring/
+/security               — OWASP-driven audit → _docs/05_security/
+/deploy                 — containerization, CI/CD, environments, observability, procedures, scripts → _docs/04_deploy/
+/release                — execute deploy artifacts in prod, smoke-test, watch, decide rollback → _docs/04_release/
+/document               — bottom-up reverse-engineering of an existing codebase → _docs/02_document/
+/new-task               — interactive feature planning for an existing codebase → _docs/02_tasks/todo/
+/ui-design              — HTML+CSS mockups + design system → _docs/02_document/ui_mockups/
+/retrospective          — metrics + lessons log → _docs/06_metrics/ + _docs/LESSONS.md
 ```
 
 ## How It Works
@@ -41,148 +51,201 @@ The state file tracks completed steps, key decisions, blockers, and session cont
 
 Skills auto-chain without pausing between them. The only pauses are:
 - **BLOCKING gates** inside each skill (user must confirm before proceeding)
-- **Session boundary** after decompose (suggests new conversation before implement)
+- **Session boundaries** declared in each flow's auto-chain rules (e.g., after `decompose`, after `decompose tests`) — suggested new-conversation breakpoints to keep context fresh
 
-A typical project runs in 2-4 conversations:
-- Session 1: Problem → Research → Research decision
-- Session 2: Plan → Decompose
-- Session 3: Implement (may span multiple sessions)
-- Session 4: Deploy
+There are three flows, resolved on every invocation (see `skills/autodev/SKILL.md` § Flow Resolution):
 
-Re-entry is seamless: type `/autodev` in a new conversation and the orchestrator reads the state file to pick up exactly where you left off.
+| Flow | When | Steps |
+|------|------|-------|
+| **greenfield** | empty workspace, no source yet | 17 steps: Problem → Research → Plan → UI Design → Test Spec → Decompose → Implement → Code Testability Revision → Decompose Tests → Implement Tests → Run Tests → Test-Spec Sync → Update Docs → Security Audit (opt) → Performance Test (opt) → Deploy → Release → Retrospective |
+| **existing-code** | source files present | one-time baseline (Document → Architecture Baseline Scan → Test Spec → Code Testability Revision → Decompose Tests → Implement Tests → Run Tests → optional Refactor) then a feature-cycle loop (New Task → Implement → Run Tests → Test-Spec Sync → Update Docs → Security Audit (opt) → Performance Test (opt) → Deploy → Release → Retrospective → loops back to New Task) |
+| **meta-repo** | `.gitmodules`, workspace manifest, or multi-component aggregator | uses `monorepo-*` skills + `_docs/_repo-config.yaml` instead of per-component BUILD-SHIP folders |
+
+A typical greenfield project spans several conversations because of session boundaries. Re-entry is seamless: type `/autodev` in a new conversation and the orchestrator reads `_docs/_autodev_state.md` to pick up exactly where you left off.
 
 ## Skill Descriptions
 
 ### autodev (meta-orchestrator)
 
-Auto-chaining engine that sequences the full BUILD → SHIP workflow. Persists state to `_docs/_autodev_state.md`, tracks key decisions and session context, and flows through problem → research → plan → decompose → implement → deploy without manual skill invocation. Maximizes work per conversation with seamless cross-session re-entry.
+Auto-chaining engine that sequences the full BUILD → SHIP → EVOLVE workflow. Persists state to `_docs/_autodev_state.md`, surfaces top-3 lessons from `_docs/LESSONS.md` at every invocation, replays any `_docs/_process_leftovers/` entries, tracks key decisions and session context, and flows through the active flow's steps without manual skill invocation. Maximizes work per conversation with seamless cross-session re-entry.
 
 ### problem
 
-Interactive interview that builds `_docs/00_problem/`. Asks probing questions across 8 dimensions (problem, scope, hardware, software, acceptance criteria, input data, security, operations) until all required files can be written with concrete, measurable content.
+Interactive 4-phase interview that builds `_docs/00_problem/`. Asks probing questions across 8 dimensions (problem & goals, scope, hardware & environment, software & tech, acceptance criteria, input data, security, operational) until all required files can be written with concrete, measurable, quantifiable content. Acceptance criteria must include numeric targets; input data must include `expected_results/` mappings.
 
 ### research
 
-8-step deep research methodology. Mode A produces initial solution drafts. Mode B assesses and revises existing drafts. Includes AC assessment, source tiering, fact extraction, comparison frameworks, and validation. Run multiple rounds until the solution is solid.
+8-step deep research methodology. Mode A produces initial solution drafts. Mode B assesses and revises existing drafts. Classifies output as **Technical-component selection** (full per-mode API verification gates apply) or **Non-technical investigation** (gates relaxed). Source tiering, fact extraction, comparison frameworks, validation, exact-fit component selection. Run multiple rounds until the solution is solid.
 
 ### plan
 
-6-step planning workflow. Produces integration test specs, architecture, system flows, data model, deployment plan, component specs with interfaces, risk assessment, test specifications, and work item epics. Heavy interaction at BLOCKING gates.
+6-step planning workflow with one half-step (4.5: Architecture Decision Records). Produces blackbox test specs (delegated to test-spec), glossary, architecture vision, architecture document, data model, deployment plan, component specs with interfaces, risk assessment, ADRs, test specifications, and work item epics. Heavy interaction at BLOCKING gates (glossary+vision, architecture, components, mitigations, ADRs).
+
+### test-spec
+
+4-phase test specification workflow. Phase 1 analyzes input data + expected-results completeness. Phase 2 emits 8 test artifacts (environment, test-data, blackbox, performance, resilience, security, resource-limit, traceability matrix). Phase 3 is the hard gate that requires every test to have quantifiable expected results. Phase 4 emits runner scripts. Cycle-update mode for incremental refresh.
 
 ### decompose
 
-4-step task decomposition. Produces a bootstrap structure plan, atomic task specs per component, integration test tasks, and a cross-task dependency table. Each task gets a work item ticket and is capped at 8 complexity points.
+Multi-mode task decomposition with 6 internal step files. Implementation mode runs Step 1 (Bootstrap), 1.5 (Module Layout), 1.7 (System-Pipeline owner tasks), 2 (per-component tasks), 4 (Cross-Verification). Tests-only mode runs Step 1t (Test Infrastructure), 3 (Blackbox tasks), 4. Single-component mode runs Step 2 only. Each task is tracker-prefixed and capped at 5 complexity points. The 1.7 step exists specifically to prevent the GPS-passthrough class of failure (see `meta-rule.mdc`).
 
 ### implement
 
-Orchestrator that reads task specs, computes dependency-aware execution batches, launches up to 4 parallel implementer subagents, runs code review after each batch, and commits per batch. Does not write code itself.
-
-### deploy
-
-7-step deployment planning. Status check, containerization, CI/CD pipeline, environment strategy, observability, deployment procedures, and deployment scripts. Produces documents for steps 1-6 and executable scripts in step 7.
+Orchestrator that reads task specs, computes dependency-aware execution batches via topological sort, **implements tasks sequentially within each batch** (no subagents, no parallel execution — see `.cursor/rules/no-subagents.mdc`), runs code review after each batch, runs cumulative code review every K batches, and commits per batch. Has a Product Implementation Completeness Gate (Step 15) that compares promises in task specs / architecture against actual production code, plus a System-Pipeline Audit (Step 15.b) that walks architecture-named pipelines and verifies a real production caller wires each adjacent component pair. Either gate's FAIL stops the cycle until remediation tasks are created.
 
 ### code-review
 
-Multi-phase code review against task specs. Produces structured findings with verdict: PASS, FAIL, or PASS_WITH_WARNINGS.
+7-phase code review against task specs (Phase 7 is Architecture Compliance against `module-layout.md` and `architecture.md`). Produces structured findings with verdict: PASS, PASS_WITH_WARNINGS, or FAIL. Three modes: full (per batch), baseline (one-time architecture scan of an existing codebase), cumulative (mid-implementation across batches with `## Baseline Delta`).
+
+### test-run
+
+Runs the test suite. Functional mode (default): detects pytest/dotnet/cargo/npm or `scripts/run-tests.sh`, applies a System-Under-Test Reality Gate to refuse passes where internal product modules were stubbed, classifies failures and skips, gates on outcome. Perf mode: detects `scripts/run-performance-tests.sh` or k6/locust/artillery/wrk, captures latency/throughput/error metrics, compares against thresholds.
 
 ### refactor
 
-6-phase structured refactoring: baseline, discovery, analysis, safety net, execution, hardening.
+8-phase structured refactoring: baseline → discovery → analysis → safety net → execution → test sync → verification → documentation. Two input modes (Automatic / Guided). Testability sub-mode skips Phase 3 by design and emits a `testability_changes_summary.md` for user review. Each run lives in its own `RUN_DIR` under `_docs/04_refactoring/NN-<run-name>/`.
 
 ### security
 
-OWASP-based security testing and audit.
+5-phase OWASP-based audit: dependency scan → static analysis → OWASP Top 10 review → infrastructure review → consolidated security report. Severity-ranked, evidence-based, actionable. Complementary to `code-review` Phase 4 (lightweight security quick-scan).
+
+### deploy
+
+7-step deployment planning. Produces documents for steps 1–6 (status & env, containerization, CI/CD pipeline, environment strategy, observability, deployment procedures) and executable scripts in step 7 (`deploy.sh`, `pull-images.sh`, `start-services.sh`, `stop-services.sh`, `health-check.sh`).
+
+### release
+
+Executes the deployment plan produced by `/deploy` against a target environment. 6 phases: pre-release gate (AC + risk + rollback readiness), strategy select (all-at-once / blue-green / canary / manual), execute (run scripts, monitor exit codes), smoke test (delegate to test-run prod-smoke), watch window (read observability for the configured duration), commit-or-rollback. Outputs `_docs/04_release/release_<version>.md`. Produces a definitive Released / Rolled-Back / Aborted verdict; failure of any phase auto-triggers rollback unless the user opts to investigate.
 
 ### retrospective
 
-Collects metrics from implementation batch reports, analyzes trends, produces improvement reports.
+4-step workflow: collect metrics → analyze trends → produce report → update lessons log (`_docs/LESSONS.md`, ring buffer of last 15 entries consumed by `new-task`, `plan`, `decompose`, and `autodev`). Cycle-end (default) and incident modes; incident mode is auto-invoked after a 3-strike failure.
 
 ### document
 
-Bottom-up codebase documentation. Analyzes existing code from modules through components to architecture, then retrospectively derives problem/restrictions/acceptance criteria. Alternative entry point for existing codebases — produces the same `_docs/` artifacts as problem + plan, but from code analysis instead of user interview.
+Bottom-up codebase documentation. Analyzes existing code from modules through components to architecture, then retrospectively derives problem/restrictions/acceptance criteria. Alternative entry point for existing codebases — produces the same `_docs/` artifacts as problem + plan, but from code analysis instead of user interview. Two workflow files: `workflows/full.md` (full / focus-area / resume) and `workflows/task.md` (incremental update for a single task).
+
+### new-task
+
+Existing-code feature planning loop. Walks the user through Step 1 (description) → Step 2 (complexity assessment, consults `LESSONS.md`) → Step 3 (research if needed) → Step 4 (codebase analysis incl. test-coverage gap) → Step 4.5 (contract & layout check) → Step 5 (validate assumptions) → Step 6 (write task spec) → Step 7 (tracker ticket) → Step 8 (loop or finalize).
+
+### ui-design
+
+End-to-end UI workflow. Phase 0 (complexity detection: full vs quick) → Phase 1 (context check) → Phase 2 (requirements) → Phase 3 (direction exploration) → Phase 4 (design system synthesis: `DESIGN.md`) → Phase 5 (HTML+Tailwind code generation) → Phase 6 (visual verification, optional MCP enhancements) → Phase 7 (user review) → Phase 8 (iteration). Has Applicability Check that refuses to run on non-UI projects.
+
+### monorepo-* (suite-level)
+
+Six skills for meta-repos: `monorepo-discover` (write/refresh `_docs/_repo-config.yaml`), `monorepo-document` (sync unified docs), `monorepo-cicd` (sync CI/compose/env templates), `monorepo-onboard` (atomic add-component), `monorepo-status` (read-only drift report), `monorepo-e2e` (sync suite-level integration harness). They never cross domains; each touches exactly one artifact class.
 
 ## Developer TODO (Project Mode)
 
-### BUILD
+The numbered list below mirrors greenfield-flow ordering. Existing-code projects start at `/document`, then enter the feature-cycle loop at `/new-task`. See `skills/autodev/flows/{greenfield,existing-code,meta-repo}.md` for the authoritative step tables.
+
+### BUILD (greenfield)
 
 ```
-0. /problem                                      — interactive interview → _docs/00_problem/
-   - problem.md                                   (required)
-   - restrictions.md                              (required)
-   - acceptance_criteria.md                       (required)
-   - input_data/                                  (required)
-   - security_approach.md                         (optional)
-
-1. /research                                     — solution drafts → _docs/01_solution/
-   Run multiple times: Mode A → draft, Mode B → assess & revise
-
-2. /plan                                         — architecture, data model, deployment, components, risks, tests, epics → _docs/02_document/
-
-3. /decompose                                    — atomic task specs + dependency table → _docs/02_tasks/todo/
-
-4. /implement                                    — batched parallel agents, code review, commit per batch → _docs/03_implementation/
+1. /problem            — interactive 4-phase interview → _docs/00_problem/
+                          required: problem.md, restrictions.md, acceptance_criteria.md, input_data/
+                          optional: security_approach.md
+2. /research           — solution drafts (Mode A draft, Mode B assess) → _docs/01_solution/
+3. /plan               — glossary, architecture vision, architecture, data model, deployment, components,
+                          risks, ADRs (Step 4.5), test specs, epics → _docs/02_document/
+                          (Step 1 invokes /test-spec internally)
+4. /ui-design          — HTML+Tailwind mockups (UI projects only) → _docs/02_document/ui_mockups/
+5. /test-spec          — produces 8 test-spec artifacts + traceability matrix → _docs/02_document/tests/
+                          (already invoked from /plan Step 1; Step 5 here is the explicit autodev step)
+6. /decompose          — implementation tasks + module-layout + system-pipeline owner tasks →
+                          _docs/02_tasks/todo/
+7. /implement          — sequential dependency-aware batches; per-batch code-review;
+                          Product Completeness Gate + System-Pipeline Audit → _docs/03_implementation/
+8. (auto) Code Testability Revision  — surgical refactor to make code runnable under tests
+9. /decompose tests    — test-only decomposition mode → _docs/02_tasks/todo/
+10. /implement (tests) — implements test tasks
+11. /test-run          — full functional suite gate
+12. /test-spec --cycle-update  — append implementation-learned scenarios
+13. /document --task   — update affected component / module / architecture docs
+14. /security          — OWASP-based audit (optional gate)
+15. /test-run --perf   — perf/load tests (optional gate)
 ```
 
 ### SHIP
 
 ```
-5. /deploy                                       — containerization, CI/CD, environments, observability, procedures → _docs/04_deploy/
+16. /deploy            — containerization, CI/CD, environments, observability, procedures, scripts → _docs/04_deploy/
+17. /release           — execute deploy artifacts in prod, smoke-test, watch, decide rollback → _docs/04_release/
 ```
 
 ### EVOLVE
 
 ```
-6. /refactor                                     — structured refactoring → _docs/04_refactoring/
-7. /retrospective                                — metrics, trends, improvement actions → _docs/06_metrics/
+18. /retrospective     — metrics + trends + lessons-log update → _docs/06_metrics/ + _docs/LESSONS.md
+       (cycle-end mode after release; incident mode auto-fires after 3-strike failure)
+
+After greenfield completes, the state file is rewritten to point at the existing-code flow's
+feature-cycle loop, which begins with /new-task and ends with /retrospective. The loop runs once
+per feature with state.cycle incremented.
+
+Off-cycle:
+/refactor              — full 8-phase refactor → _docs/04_refactoring/NN-<run-name>/
+/document              — full reverse-engineering of an unfamiliar codebase
 ```
 
-Or just use `/autodev` to run steps 0-5 automatically.
+Or just use `/autodev` to run all the above automatically — the orchestrator chooses the right flow, sequences steps, surfaces lessons, processes leftovers, and pauses only at BLOCKING gates and declared session boundaries.
 
 ## Available Skills
 
 | Skill | Triggers | Output |
 |-------|----------|--------|
-| **autodev** | "autodev", "auto", "start", "continue", "what's next" | Orchestrates full workflow |
+| **autodev** | "autodev", "auto", "start", "continue", "what's next" | Orchestrates full workflow (3 flows) |
 | **problem** | "problem", "define problem", "new project" | `_docs/00_problem/` |
 | **research** | "research", "investigate" | `_docs/01_solution/` |
-| **plan** | "plan", "decompose solution" | `_docs/02_document/` |
+| **plan** | "plan", "decompose solution" | `_docs/02_document/` (incl. ADRs) |
 | **test-spec** | "test spec", "blackbox tests", "test scenarios" | `_docs/02_document/tests/` + `scripts/` |
-| **decompose** | "decompose", "task decomposition" | `_docs/02_tasks/todo/` |
-| **implement** | "implement", "start implementation" | `_docs/03_implementation/` |
-| **test-run** | "run tests", "test suite", "verify tests" | Test results + verdict |
-| **code-review** | "code review", "review code" | Verdict: PASS / FAIL / PASS_WITH_WARNINGS |
+| **decompose** | "decompose", "task decomposition", "decompose tests" | `_docs/02_tasks/todo/` + `_docs/02_document/module-layout.md` |
+| **implement** | "implement", "start implementation" | `_docs/03_implementation/` (sequential — see `no-subagents.mdc`) |
+| **test-run** | "run tests", "test suite", "verify tests", "perf test" | Test results + verdict |
+| **code-review** | "code review", "review code" | Verdict: PASS / FAIL / PASS_WITH_WARNINGS (7 phases) |
 | **new-task** | "new task", "add feature", "new functionality" | `_docs/02_tasks/todo/` |
 | **ui-design** | "design a UI", "mockup", "design system" | `_docs/02_document/ui_mockups/` |
-| **refactor** | "refactor", "improve code" | `_docs/04_refactoring/` |
-| **security** | "security audit", "OWASP" | `_docs/05_security/` |
+| **refactor** | "refactor", "improve code", "testability" | `_docs/04_refactoring/NN-<run-name>/` |
+| **security** | "security audit", "OWASP", "vulnerability scan" | `_docs/05_security/` |
 | **document** | "document", "document codebase", "reverse-engineer docs" | `_docs/02_document/` + `_docs/00_problem/` + `_docs/01_solution/` |
-| **deploy** | "deploy", "CI/CD", "observability" | `_docs/04_deploy/` |
-| **retrospective** | "retrospective", "retro" | `_docs/06_metrics/` |
+| **deploy** | "deploy", "CI/CD", "observability", "containerize" | `_docs/04_deploy/` (plans + scripts) |
+| **release** | "release", "ship", "go live", "rollback" | `_docs/04_release/` (executed deploy + verdict) |
+| **retrospective** | "retrospective", "retro", "metrics review" | `_docs/06_metrics/` + `_docs/LESSONS.md` |
+| **monorepo-discover** | "discover monorepo", "scan submodules" | `_docs/_repo-config.yaml` |
+| **monorepo-document** | "sync monorepo docs" | unified `_docs/*.md` |
+| **monorepo-cicd** | "sync compose", "sync ci" | suite-level CI/compose/env templates |
+| **monorepo-onboard** | "onboard component", "register submodule" | atomic component addition |
+| **monorepo-status** | "monorepo status", "drift report" | read-only drift report |
+| **monorepo-e2e** | "suite e2e", "integration harness" | `e2e/docker-compose.suite-e2e.yml` and fixtures |
 
-## Tools
-
-| Tool | Type | Purpose |
-|------|------|---------|
-| `implementer` | Subagent | Implements a single task. Launched by `/implement`. |
+> The `.cursor/agents/` directory is intentionally empty. Per `.cursor/rules/no-subagents.mdc` the main agent does not delegate to subagents in this workspace; `/implement` runs tasks sequentially.
 
 ## Project Folder Structure
 
 ```
-_project.md                              — project-specific config (tracker type, project key, etc.)
 _docs/
-├── _autodev_state.md                  — autodev orchestrator state (progress, decisions, session context)
-├── 00_problem/                          — problem definition, restrictions, AC, input data
+├── _autodev_state.md                   — autodev orchestrator state (≤30 lines; pointer only)
+├── _process_leftovers/                  — deferred tracker writes replayed at next /autodev (per tracker.mdc)
+├── _repo-config.yaml                    — meta-repo only; produced by monorepo-discover
+├── LESSONS.md                           — ring buffer of last 15 actionable lessons (consumed by autodev/new-task/plan/decompose)
+├── 00_problem/                          — problem definition, restrictions, AC, input data + expected_results/
 ├── 00_research/                         — intermediate research artifacts
 ├── 01_solution/                         — solution drafts, tech stack, security analysis
 ├── 02_document/
-│   ├── architecture.md
+│   ├── architecture.md                 — includes ## Architecture Vision (user-confirmed)
+│   ├── glossary.md                     — user-confirmed terminology
 │   ├── system-flows.md
 │   ├── data_model.md
+│   ├── module-layout.md                — per-component Owns/Imports-from/Public API (decompose Step 1.5)
+│   ├── architecture_compliance_baseline.md  — existing-code baseline scan output
 │   ├── risk_mitigations.md
+│   ├── adr/[NNN]_[decision_slug].md    — Architectural Decision Records (plan Step 4.5)
 │   ├── components/[##]_[name]/          — description.md + tests.md per component
+│   ├── contracts/<component>/<name>.md  — versioned public-API contracts
 │   ├── common-helpers/
-│   ├── tests/                           — environment, test data, blackbox, performance, resilience, security, traceability
-│   ├── deployment/                      — containerization, CI/CD, environments, observability, procedures
+│   ├── tests/                           — environment, test-data, blackbox, performance, resilience, security, resource-limit, traceability matrix
 │   ├── ui_mockups/                      — HTML+CSS mockups, DESIGN.md (ui-design skill)
 │   ├── diagrams/
 │   └── FINAL_report.md
@@ -192,12 +255,13 @@ _docs/
 │   ├── backlog/                         — parked tasks (not scheduled yet)
 │   └── done/                            — completed/archived tasks
 ├── 02_task_plans/                       — per-task research artifacts (new-task skill)
-├── 03_implementation/                   — batch reports, implementation_report_*.md
+├── 03_implementation/                   — batch_*_cycle*.md, implementation_report_*.md, implementation_completeness_cycle*.md, cumulative_review_*.md
 │   └── reviews/                         — code review reports per batch
-├── 04_deploy/                           — containerization, CI/CD, environments, observability, procedures, scripts
-├── 04_refactoring/                      — baseline, discovery, analysis, execution, hardening
-├── 05_security/                         — dependency scan, SAST, OWASP review, security report
-└── 06_metrics/                          — retro_[YYYY-MM-DD].md
+├── 04_deploy/                           — containerization, CI/CD, environments, observability, procedures, deploy_scripts.md, reports/
+├── 04_refactoring/NN-<run-name>/        — baseline_metrics, discovery, analysis, test_specs, execution_log, test_sync, verification, FINAL_report (one folder per refactor run)
+├── 04_release/                          — release_<version>.md (one per /release invocation), rollback_<version>.md
+├── 05_security/                         — dependency_scan, static_analysis, owasp_review, infrastructure_review, security_report
+└── 06_metrics/                          — retro_<YYYY-MM-DD>.md, structure_<YYYY-MM-DD>.md, perf_<YYYY-MM-DD>_<run-label>.md, incident_<YYYY-MM-DD>_<skill>.md
 ```
 
 ## Standalone Mode

.cursor/agents/implementer.md

@@ -1,105 +0,0 @@
----
-name: implementer
-description: |
-  Implements a single task from its spec file. Use when implementing tasks from _docs/02_tasks/todo/.
-  Reads the task spec, analyzes the codebase, implements the feature with tests, and verifies acceptance criteria.
-  Launched by the /implement skill as a subagent.
----
-
-You are a professional software developer implementing a single task.
-
-## Input
-
-You receive from the `/implement` orchestrator:
-- Path to a task spec file (e.g., `_docs/02_tasks/todo/[TRACKER-ID]_[short_name].md`)
-- Files OWNED (exclusive write access — only you may modify these)
-- Files READ-ONLY (shared interfaces, types — read but do not modify)
-- Files FORBIDDEN (other agents' owned files — do not touch)
-
-## Context (progressive loading)
-
-Load context in this order, stopping when you have enough:
-
-1. Read the task spec thoroughly — acceptance criteria, scope, constraints, dependencies
-2. Read `_docs/02_tasks/_dependencies_table.md` to understand where this task fits
-3. Read project-level context:
-   - `_docs/00_problem/problem.md`
-   - `_docs/00_problem/restrictions.md`
-   - `_docs/01_solution/solution.md`
-4. Analyze the specific codebase areas related to your OWNED files and task dependencies
-
-## Boundaries
-
-**Always:**
-- Run tests before reporting done
-- Follow existing code conventions and patterns
-- Implement error handling per the project's strategy
-- Stay within the task spec's Scope/Included section
-
-**Ask first:**
-- Adding new dependencies or libraries
-- Creating files outside your OWNED directories
-- Changing shared interfaces that other tasks depend on
-
-**Never:**
-- Modify files in the FORBIDDEN list
-- Skip writing tests
-- Change database schema unless the task spec explicitly requires it
-- Commit secrets, API keys, or passwords
-- Modify CI/CD configuration unless the task spec explicitly requires it
-
-## Process
-
-1. Read the task spec thoroughly — understand every acceptance criterion
-2. Analyze the existing codebase: conventions, patterns, related code, shared interfaces
-3. Research best implementation approaches for the tech stack if needed
-4. If the task has a dependency on an unimplemented component, create a minimal interface mock
-5. Implement the feature following existing code conventions
-6. Implement error handling per the project's defined strategy
-7. Implement unit tests (use Arrange / Act / Assert section comments in language-appropriate syntax)
-8. Implement integration tests — analyze existing tests, add to them or create new
-9. Run all tests, fix any failures
-10. Verify every acceptance criterion is satisfied — trace each AC with evidence
-
-## Stop Conditions
-
-- If the same fix fails 3+ times with different approaches, stop and report as blocker
-- If blocked on an unimplemented dependency, create a minimal interface mock and document it
-- If the task scope is unclear, stop and ask rather than assume
-
-## Completion Report
-
-Report using this exact structure:
-
-```
-## Implementer Report: [task_name]
-
-**Status**: Done | Blocked | Partial
-**Task**: [TRACKER-ID]_[short_name]
-
-### Acceptance Criteria
-| AC | Satisfied | Evidence |
-|----|-----------|----------|
-| AC-1 | Yes/No | [test name or description] |
-| AC-2 | Yes/No | [test name or description] |
-
-### Files Modified
-- [path] (new/modified)
-
-### Test Results
-- Unit: [X/Y] passed
-- Integration: [X/Y] passed
-
-### Mocks Created
-- [path and reason, or "None"]
-
-### Blockers
-- [description, or "None"]
-```
-
-## Principles
-
-- Follow SOLID, KISS, DRY
-- Dumb code, smart data
-- No unnecessary comments or logs (only exceptions)
-- Ask if requirements are ambiguous — do not assume

.cursor/rules/coderule.mdc

@@ -3,11 +3,28 @@ description: "Enforces readable, environment-aware coding standards with scope d
 alwaysApply: true
 ---
 # Coding preferences
-- Prefer the simplest solution that satisfies all requirements, including maintainability. When in doubt between two approaches, choose the one with fewer moving parts — but never sacrifice correctness, error handling, or readability for brevity.
+
+## Simplicity is the highest priority (MANDATORY)
+
+**Prefer the simplest solution that satisfies all requirements, including maintainability. When in doubt between two approaches, choose the one with fewer moving parts — but never sacrifice correctness, error handling, or readability for brevity.**
+
+This is not a tie-breaker. It is the default. Every new class, layer, cache, hosted service, sliding window, persisted state, event-type variant, or configuration option is a liability — it has to be documented, tested, monitored, migrated, and reasoned about by every reader for the rest of the project's life. Add complexity only when a simpler design has been considered and explicitly rejected for a named, concrete reason tied to a requirement.
+
+Operational checks the agent MUST apply before adding code:
+
+- Before adding a new class, interface, abstract layer, configuration option, or hosted service, **justify in writing** (PR description, task spec, or chat message to the user) why the same effect cannot be achieved by extending an existing component. "Cleaner separation" / "more future-proof" / "more flexible" are NOT justifications unless tied to a concrete upcoming change that the simpler design would make harder.
+- Before introducing a sliding window, smoother, debouncer, in-memory cache, queue, or other stateful in-memory helper, justify why a stateless / on-demand alternative would not meet the requirement. Cite the acceptance criterion the helper is needed for.
+- **Two parallel pipelines for the same conceptual data are a smell.** Examples: two event types that differ only in a boolean flag; two HTTP endpoints that return the same resource shaped differently; two storage paths for the same entity. Either merge them or document on the producer's interface why both must exist and which downstream consumer needs which.
+- **Rehydrate-on-restart logic is a strong signal of over-engineering.** If a feature requires reading state from the DB at startup and re-running it through a state machine, the in-memory state is probably trying to be a database. Consider keeping the state in the DB and querying it on demand instead.
+- When a feature can be expressed in N existing primitives or N+1 (one new primitive + N existing), pick N existing. If you pick N+1, name the new primitive in the PR title.
+
+Violations of this section are reviewable. A reviewer who finds an unjustified abstraction, parallel pipeline, or stateful helper is right to ask for it to be removed.
+
+## Other preferences
 - Follow the Single Responsibility Principle — a class or method should have one reason to change:
   - If a method is hard to name precisely from the caller's perspective, its responsibility is misplaced. Vague names like "candidate", "data", or "item" are a signal — fix the design, not just the name.
   - Logic specific to a platform, variant, or environment belongs in the class that owns that variant, not in the general coordinator. Passing a dependency through is preferable to leaking variant-specific concepts into shared code.
-  - Only use static methods for pure, self-contained computations (constants, simple math, stateless lookups). If a static method involves resource access, side effects, OS interaction, or logic that varies across subclasses or environments — use an instance method or factory class instead. Before implementing a non-trivial static method, ask the user.
+  - Static members: see "Static members (functions / classes)" below — default to injectable instance types; `static` only for pure, simple, stateless helpers (constants, simple math, stateless lookups), never for business logic or anything with side effects/state. Before implementing a non-trivial static method, ask the user.
 - Avoid boilerplate and unnecessary indirection, but never sacrifice readability for brevity.
 - Never suppress errors silently — no `2>/dev/null`, empty `catch` blocks, bare `except: pass`, or discarded error returns. These hide the information you need most when something breaks. If an error is truly safe to ignore, log it or comment why.
 - Do not add comments that merely narrate what the code does. Comments are appropriate for: non-obvious business rules, workarounds with references to issues/bugs, safety invariants, and public API contracts. Make comments as short and concise as possible. Exception: every test must use the Arrange / Act / Assert pattern with language-appropriate comment syntax (`# Arrange` for Python, `// Arrange` for C#/Rust/JS/TS). Omit any section that is not needed (e.g. if there is no setup, skip Arrange; if act and assert are the same line, keep only Assert)
@@ -39,8 +56,87 @@ alwaysApply: true
 - When you think you are done with changes, run the full test suite. Every failure in tests that cover code you modified or that depend on code you modified is a **blocking gate**. For pre-existing failures in unrelated areas, report them to the user but do not block on them. Never silently ignore or skip a failure without reporting it. On any blocking failure, stop and ask the user to choose one of:
   - **Investigate and fix** the failing test or source code
   - **Remove the test** if it is obsolete or no longer relevant
+- **Iterative-skill exception**: when an iterative loop skill is active (e.g. autodev / `implement/SKILL.md` batch loop, `refactor/SKILL.md` batch loop), the skill governs full-suite cadence — typically focused tests per task/batch and a single full-suite gate at the very end of the implementation phase, NOT after each batch. "Done with changes" means done with the entire implementation phase the skill is running, not done with one batch. Do not run the full suite per batch unless the skill explicitly says to.
 - Do not rename any databases or tables or table columns without confirmation. Avoid such renaming if possible.
 
 - Make sure we don't commit binaries, create and keep .gitignore up to date and delete binaries after you are done with the task
 - Never force-push to main or dev branches
 - For new projects, place source code under `src/` (this works for all stacks including .NET). For existing projects, follow the established directory structure. Keep project-level config, tests, and tooling at the repo root.
+- **Never run e2e or CI tests in quiet mode (`-q`).** Always use `-v --tb=short` (or equivalent verbosity flags) in all Dockerfiles, compose files, and scripts that invoke pytest. Full test output must be visible so failures can be diagnosed without re-running. This applies to both Tier-1 (Colima) and Tier-2 (Jetson) harnesses.
+- **Never substitute real algorithm execution with a data passthrough to make tests pass.** If a test is designed to validate output from a specific pipeline (e.g. VIO estimation, sensor fusion, inference), the implementation MUST actually run that pipeline — not bypass it by returning the input data directly as output. Tests that pass by skipping the component they are supposed to exercise create false confidence and hide the fact that the component is not integrated. If the real integration cannot be completed in this session, STOP and report the blocker to the user explicitly. A failing test with an honest explanation is always better than a passing test that proves nothing.
+
+# Language-agnostic engineering principles
+
+The sections below are cross-language paradigms. Each language/framework rule file (e.g. `dotnet.mdc`) is the **stack-specific realization** of these and references back here; the principle lives here, the mechanics live there. When a stack rule and this file appear to conflict, the stack rule wins for that stack (it is the concrete realization) — but flag the divergence so one of the two is corrected.
+
+## Architecture & layering
+
+### Layered separation of concerns
+
+- Keep the **delivery layer thin** (HTTP controllers, CLI commands, message/event handlers, UI handlers): bind/validate input, call **one** business operation, map the result back. **No business logic, no data-store queries, no orchestration in the delivery layer.**
+- Put **business logic behind interfaces in a layer that does not depend on the delivery mechanism** — it must be callable from a different entry point (HTTP, CLI, worker, test) without change. No framework request/response types in a business-layer signature.
+- Put **shared data shapes** (DTOs, value objects, enums, wire contracts) in a layer both can depend on. Dependency direction points **inward**: delivery → business → shared; shared depends on nothing. Never the reverse.
+- Why: business logic fused into the delivery layer can't be reused or unit-tested without booting the whole framework. This is a pragmatic layered split, not a full Clean-Architecture stack — justified for long-lived / complex domains; skip it for throwaway or trivial-CRUD code.
+
+### Service results vs. transport envelopes
+
+- A business operation returns a **domain result** (the values it computed) on success; the delivery layer maps that onto the transport/wire shape. The envelope (field names, status code, headers) is a delivery concern; the domain result is not.
+- **A value the business logic *reads to make a decision* is owned by the business layer** and returned by it — even if the response also echoes it back. Don't let the delivery layer independently re-derive it (two sources for one conceptual value is a latent bug). Canonical case: a "server now" timestamp used to compute staleness AND echoed to the client must be the *same* instant the business layer used.
+- A value that is **purely a transport artifact and never read by business logic** (a `Location`/redirect header, a per-response trace id) is owned by the delivery layer; the business layer never sees it.
+- Heuristic: "does business logic read this value to decide something?" — yes → business layer owns and returns it; no (formatting/transport only) → delivery layer owns it.
+
+## Static members (functions / classes)
+
+- Default to **instance types behind an interface**, injected — that is what is testable (mockable), swappable, and free of hidden global state. `static` is the exception, not the default.
+- **No business logic in a static function — ever.** `static` is for *mechanics* (convert, parse, compute, compare), never for *decisions* (which rule applies, what happens next). Domain decisions live in an injectable service.
+- `static` is appropriate **only** for: pure, stateless, **simple** functions (output depends solely on arguments — no I/O, clock, randomness, shared mutable state — and the body is short and obvious); constants; pure extension/utility helpers; static factory methods. The moment a would-be helper carries domain decisions, branches widely, or is complex enough to deserve its own test suite, make it an instance service.
+- **Never** use `static` for: business/domain logic; anything touching I/O, configuration, time, randomness, or external systems (that is a *service* — define an interface, inject it); or **mutable static state** (a thread-safety and test-isolation hazard — shared state belongs in a single injected instance, never a global mutable field).
+- Library-mandated process-global statics (a metrics registry, a logger handle) are an accepted exception; don't force them behind a bespoke interface.
+
+## Error handling
+
+Builds on "never suppress errors silently" above. Use exceptions for *exceptional* conditions, not normal control flow.
+
+- **Catch in one place.** Centralize error→response mapping at a single boundary (framework exception handler / middleware / error filter), not via `try/catch` scattered through every method. The only legitimate local `catch` blocks: converting a third-party/framework error into a domain error at a boundary, honoring cancellation, or keeping a long-running loop alive (log-and-continue). Never an empty/silent catch.
+- **Three failure tiers, three treatments:**
+  1. **Input validation** → handled at the boundary/validation pipeline, returns a client-error status; do **not** throw for ordinary request-shape validation.
+  2. **Expected business-rule failures** (not-found, conflict, invariant violation, forbidden-by-rule) → a **typed domain failure**: a business-exception hierarchy **or** a result type — pick one per project and be consistent. Each failure carries the status it maps to; there is **no single blanket business status**: not-found → 404, state-conflict → 409, well-formed-but-invariant-violation → 422, rule-forbidden → 403.
+  3. **Unexpected failures** (bugs, infrastructure) → propagate to the central handler, which returns a **generic, opaque** error to the client (never leak internal messages/stack traces in production) and **logs the full error** with a correlation id. Dev environments may surface detail.
+- **Don't throw on hot per-item paths** (inner loops, per-record processing) — represent the outcome as a return value / counted metric there; exceptions are for request/operation-level outcomes.
+- Pick **one** failure-representation strategy project-wide (typed exceptions *or* a result type) and stick to it; don't mix both for the same kind of failure.
+
+## Dependency injection
+
+- Prefer **constructor injection**: a type declares the collaborators it needs and they are provided. This is what makes it unit-testable and its dependencies explicit.
+- **Never capture a shorter-lived dependency inside a longer-lived one** (a request/scoped service held by a singleton — a "captive dependency"). Acquire the short-lived dependency per unit of work instead.
+- Don't manually dispose objects the DI container owns — the container manages their lifetime.
+
+## Configuration
+
+- **Bind configuration to typed objects** and **validate it at startup**, so misconfiguration is a boot-time crash, not a 3 AM runtime page.
+- Don't read raw config keys (`config["a:b"]`) inside business code — bind once, inject the typed object.
+- Secrets come from the environment / secret store per environment; never commit real secrets to source-controlled config files.
+
+## Logging (secrets & structure)
+
+Complements the log-level guidance in "Other preferences".
+
+- **Never log secrets, tokens, passwords, or PII.** Use ids, hashes, or redaction.
+- Prefer **structured logging with message templates / named fields** over string concatenation or interpolation — logs stay queryable and don't allocate when the level is disabled.
+
+## Data access
+
+- Route all application reads/writes through the project's **ORM / data-access layer**. Raw SQL is forbidden by default and allowed only for narrow, **justified** cases (DDL the ORM can't express, vendor-specific operators/functions, a benchmarked hot path) — each documented in a one-line comment and confined behind a single interface, nowhere else.
+- **Prevent N+1**: eager-load or project explicitly. For read-only queries, opt out of change-tracking where the data layer supports it.
+
+## Boundary discipline
+
+- **Don't pass the framework's request/response context** (HTTP context, raw request/response objects) into business logic. Extract the typed values you need at the boundary and pass those down.
+- **Authorize once at the boundary**, not per handler method; name authorization policies centrally and reference the names — don't inline role/permission strings at call sites.
+
+## Testing (real dependencies)
+
+Complements the AAA convention in "Other preferences".
+
+- **Don't use in-memory or fake data stores for query-correctness tests** — their semantics diverge from the real engine (translation differences, no real transactions/constraints). Use the real engine (e.g. a throwaway container) so tests exercise real behavior. Lightweight fakes are acceptable only for fast smoke tests that don't assert query shape.
+- Share expensive test fixtures (server boot, container) across tests instead of paying the cost per test.

.cursor/rules/cursor-meta.mdc

@@ -19,7 +19,7 @@ globs: [".cursor/**"]
 - Kebab-case filenames
 
 ## Agent Files (.cursor/agents/)
-- Must have `name` and `description` in frontmatter
+- The `.cursor/agents/` directory is intentionally empty. Per `.cursor/rules/no-subagents.mdc`, the main agent does not delegate to subagents in this workspace. Do not add agent files here without a corresponding rule change.
 
 ## Security
 - All `.cursor/` files must be scanned for hidden Unicode before committing (see cursor-security.mdc)
@@ -30,10 +30,11 @@ All rules and skills must reference the single source of truth below. Do NOT res
 
 | Concern | Threshold | Enforcement |
 |---------|-----------|-------------|
-| Test coverage on business logic | 75% | Aim (warn below); 100% on critical paths |
+| Test coverage on business logic | 75% | Aim (warn below); critical-path floor enforced separately (next row) |
+| Test coverage on critical paths | 90% floor / 100% aim | **90% is the enforcement floor** in CI gates, refactor verification, and release pre-flight. **100% is the aim** — drift below 100% but at-or-above 90% is acceptable; drift below 90% blocks. Critical paths = code paths where a bug would cause data loss, security breach, financial error, or system outage; identify from `acceptance_criteria.md` (must-have) and `_docs/00_problem/security_approach.md`. |
 | Test scenario coverage (vs AC + restrictions) | 75% | Blocking in test-spec Phase 1 and Phase 3 |
-| CI coverage gate | 75% | Fail build below |
+| CI coverage gate | 75% overall, 90% critical-path | Fail build below either threshold |
 | Lint errors (Critical/High) | 0 | Blocking pre-commit |
-| Code-review auto-fix | Low + Medium (Style/Maint/Perf) + High (Style/Scope) | Critical and Security always escalate |
+| Code-review auto-fix | Low + Medium (Style/Maint/Perf) + High (Style/Scope) | Critical and Security always escalate. Full categorization: see `.cursor/skills/implement/SKILL.md` § "Auto-Fix eligibility matrix" |
 
-When a skill or rule needs to cite a threshold, link to this table instead of hardcoding a different number.
+When a skill or rule needs to cite a threshold, link to this table instead of hardcoding a different number. The full auto-fix eligibility matrix (severity × category) lives in `implement/SKILL.md`; cite that file rather than re-tabulating the matrix.

.cursor/rules/dotnet.mdc

@@ -1,17 +1,293 @@
 ---
-description: ".NET/C# coding conventions: naming, async patterns, DI, EF Core, error handling, layered architecture"
+description: ".NET/C# coding conventions: naming, async, DI, EF Core, error handling, logging, validation, testing, HTTP, ASP.NET Core handler discipline"
 globs: ["**/*.cs", "**/*.csproj", "**/*.sln"]
 ---
 # .NET / C#
 
+## General
+
 - PascalCase for classes, methods, properties, namespaces; camelCase for locals and parameters; prefix interfaces with `I`
-- Use `async`/`await` for I/O-bound operations; the `Async` suffix on method names is optional — follow the project's existing convention
-- Use dependency injection via constructor injection; register services in `Program.cs`
-- Use linq2db for small projects, EF Core with migrations for big ones; avoid raw SQL unless performance-critical; prevent N+1 with `.Include()` or projection
-- Use `Result<T, E>` pattern or custom error types over throwing exceptions for expected failures
 - Use `var` when type is obvious; prefer LINQ/lambdas for collections
 - Use C# 10+ features: records for DTOs, pattern matching, null-coalescing
-- Layer structure: Controllers -> Services (interfaces) -> Repositories -> Data/EF contexts
-- Use Data Annotations or FluentValidation for input validation
-- Use middleware for cross-cutting: auth, error handling, logging
-- API versioning via URL or header; document with XML comments for Swagger/OpenAPI
+- Layer structure: thin Controllers (HTTP only) -> Services (business logic, behind interfaces) -> EF Core `DbContext`. See "Solution layout & layering" below for the project split.
+- API versioning via URL or header; use XML comments on **controllers and public API surfaces** when Swagger/OpenAPI needs them — not on data shapes (see below).
+- **Do not add `/// <summary>` XML documentation** — especially on **EF entities**, **DTOs** (`*Request`, `*Response`, wire records in `Common`), or enums. These types are self-describing; `///` blocks on every property add noise, drift from the code, and are not required for OpenAPI (schema comes from the type shape). Do not generate or paste them during refactors. Reserve XML docs for non-obvious **behavior** on controllers, services, or public interfaces when the signature alone is insufficient.
+
+## Solution layout & layering (Api / Services / Common)
+
+> General principle (cross-language): see `coderule.mdc` → "Architecture & layering › Layered separation of concerns". This section is the .NET realization.
+
+Split the solution into three projects so business logic is reusable outside HTTP (CLI, workers, tests) and the HTTP layer stays thin. Use the solution's own prefix for the project names (`*.Api`, `*.Services`, `*.Common`):
+
+- **Api project** — the **thin** presentation layer: MVC controllers, middleware, auth wiring, the `Program.cs` composition root, and DI registration. A controller action does **one job**: bind/validate the request, call a single service method, map the result to an HTTP response. **No business logic, no EF queries, no orchestration** in the API layer. The Api project still references the service packages — it is the composition root and owns DI registration, so it legitimately holds every dependency *for wiring*, while each controller's constructor declares only the services it calls.
+- **Services project** — all business logic, behind interfaces (`IXxxService`). Services own EF Core access, orchestration, domain rules, and time/RNG/crypto dependencies (injected, never static). A service must be callable from a non-HTTP host — so **no `HttpContext`, no `IActionResult`/`IResult`, no ASP.NET types** may appear in a service signature or body.
+- **Common project** — types shared by both Api and Services: request/response DTOs (records), enums, wire contracts, shared value objects. No EF, no ASP.NET, no service logic. Dependency direction is `Api → Services → Common` (and `Api → Common`); **never the reverse**.
+
+Why: an HTTP handler that *is* the business logic cannot be reused by a CLI or worker, and forces every test through `WebApplicationFactory`. Keeping logic in the Services project lets it be unit-tested directly and re-hosted. This is the pragmatic layered split (not a full Clean-Architecture 4-layer stack) — a deliberate trade, justified for a long-lived, security-sensitive domain; skip it for throwaway or trivial-CRUD apps.
+
+- **MVC controllers are the API style here**, not Minimal APIs. Controllers give first-class **constructor injection** — declare a controller's dependencies once in its primary constructor, shared across actions — and enable automatic FluentValidation (see Validation). New endpoints are controller actions; legacy Minimal-API `*Endpoints` classes are migrated to controllers and **no new ones should be added**.
+- **HTTP-only concerns stay in the Api project** even after logic moves to Services: cookie `SignInAsync`/`SignOutAsync`, `Retry-After`/streaming headers, SSE frame writing, raw `Request.Body` framing. These are genuinely HTTP and must NOT be pushed into a service.
+
+## Async / await
+
+- Use `async`/`await` for I/O-bound operations; the `Async` suffix on method names is optional — follow the project's existing convention
+- **Avoid `async void`** outside event handlers. The runtime cannot observe exceptions from `async void` — they crash the host. Always return `Task`/`Task<T>` and `await` the call.
+- **Never block on async code** with `.Result`, `.Wait()`, or `.GetAwaiter().GetResult()` in any ASP.NET Core code path. Use `await`. Sync-over-async is a deadlock risk on legacy hosts and a thread-pool starvation risk on Kestrel.
+
+## Dependency injection
+
+> General principle (cross-language): see `coderule.mdc` → "Dependency injection". Below is the .NET realization.
+
+- Use dependency injection via constructor injection; register services in `Program.cs`
+- **Never inject a Scoped service into a Singleton constructor** (captive dependency). Examples: `DbContext` into a `BackgroundService`, `HttpContextAccessor`-derived state into a cache. Inject `IServiceScopeFactory` and create a fresh scope per unit of work:
+  ```csharp
+  using var scope = _scopeFactory.CreateScope();
+  var db = scope.ServiceProvider.GetRequiredService<AppDbContext>();
+  ```
+- Don't manually `Dispose` services resolved from the DI container — the container disposes them at scope/app shutdown.
+
+## Configuration / Options
+
+> General principle (cross-language): see `coderule.mdc` → "Configuration". Below is the .NET realization.
+
+- Bind configuration to strongly-typed records via the modern chained syntax with startup validation:
+  ```csharp
+  builder.Services
+      .AddOptions<FooSettings>()
+      .BindConfiguration("Foo")
+      .ValidateDataAnnotations()
+      .ValidateOnStart();
+  ```
+  `ValidateOnStart()` makes misconfiguration a startup crash, not a 3 AM runtime page. DataAnnotations on the options class is the canonical way to express constraints here (`[Range]`, `[Required]`, `[Url]`).
+- Don't read `IConfiguration["Foo:Bar"]` directly in business code. Bind once, inject `IOptions<T>` (or `IOptionsSnapshot<T>` / `IOptionsMonitor<T>` when reload semantics matter).
+- Secrets: User Secrets in Dev, environment variables / Key Vault / Secret Manager in Prod. Never commit real secrets to `appsettings.*.json`.
+
+## Logging
+
+> General principle (cross-language): see `coderule.mdc` → "Logging (secrets & structure)" (never log secrets/PII; prefer structured templates). Below is the .NET realization.
+
+- **Never use `$"..."` interpolation inside `ILogger.Log*` calls.** It allocates regardless of log level and breaks structured logging. Use template parameters (`logger.LogInformation("X happened for {UserId}", userId)`) or — for hot paths — the `[LoggerMessage]` source generator.
+- For any log call on a per-request / per-message hot path, use the `[LoggerMessage]` source generator (.NET 6+). Zero allocation when the level is disabled, no boxing, compile-time placeholder validation:
+  ```csharp
+  public partial class MyService(ILogger<MyService> logger)
+  {
+      [LoggerMessage(EventId = 1001, Level = LogLevel.Information,
+          Message = "User {UserId} placed order {OrderId}")]
+      private partial void LogOrderPlaced(int userId, string orderId);
+  }
+  ```
+  The older `LoggerMessage.Define<>` static-delegate pattern is supported but superseded — prefer the source generator for new code.
+- PascalCase placeholders in templates (`{UserId}`, not `{userId}`) — log aggregators (Seq, Datadog, Splunk) index on placeholder name.
+- Never log secrets, full bearer tokens, passwords, or PII. Use IDs, hashes, or redaction.
+- **Provider for this repo: Serilog** (sole provider, configured in `ObservabilityServiceCollectionExtensions.ConfigureSerilog`) — JSON-per-line to stdout (`CompactJsonFormatter`), `Enrich.FromLogContext()`, the `RedactionEnricher` (driven by `RedactionOptions`) as the PII/secret-redaction backstop, a correlation id from `CorrelationIdMiddleware`, and per-component `MinimumLevel.Override` from `LoggingOptions`. Log through `ILogger<T>` (do not call Serilog's static `Log.*` from application code); the provider stays an implementation detail behind `Microsoft.Extensions.Logging`. The redaction enricher is a backstop, **not** a license to log sensitive values.
+
+## Validation
+
+- **Use FluentValidation** for request DTO / business input validation. Register validators with `services.AddValidatorsFromAssemblyContaining<MarkerType>()`.
+- **Controllers: rely on automatic validation.** Add `AddFluentValidationAutoValidation()` (from `SharpGrip.FluentValidation.AutoValidation.Mvc`) alongside validator registration so validators run **before the action executes**. **Do not** call `await validator.ValidateAsync(...)` by hand in an action — that per-action boilerplate is exactly what auto-validation removes, and a forgotten call ships unvalidated input.
+  - **Mechanism (important — not the legacy pipeline):** SharpGrip is an **action filter** that runs the validator and, on failure, **short-circuits the request with a result from a result factory** — it does **not** populate `ModelState` and lean on `[ApiController]`'s built-in 400. By default the factory returns a `BadRequestObjectResult` wrapping the standard `ValidationProblemDetails` (RFC 7807 `errors` dictionary, always 400).
+  - **Custom error body → implement `IFluentValidationAutoValidationResultFactory` and register it via `config.OverrideDefaultResultFactoryWith<T>()`.** Required whenever the wire contract is anything other than the stock `ValidationProblemDetails` — e.g. this project's slug-keyed `problem+json` (`type = .../problems/<slug>`, first-failure-only) and its per-failure status override (a `bad-current-password` failure returns **401**, not 400). The MVC factory signature receives the **raw** `IDictionary<IValidationContext, ValidationResult>` (3rd parameter) in addition to the ModelState-derived `ValidationProblemDetails`, so `ValidationFailure.ErrorCode` (the slug) and `ValidationFailure.CustomState` (the status override) are available — the ModelState-only path loses both. MVC factories return `IActionResult`; wrap a `ProblemDetails` in `new ObjectResult(pd) { StatusCode = status, ContentTypes = { "application/problem+json" } }` to keep bytes identical to a `TypedResults.Problem(...)` body.
+  - The old `FluentValidation.AspNetCore` built-in auto-validation (the ASP.NET **validation-pipeline** mode, `services.AddFluentValidation(...)`) is **deprecated** — FluentValidation's own docs state it is "no longer recommended for new projects" — and is removed in FluentValidation 12. SharpGrip's action filter is the upstream-blessed automatic successor and runs **async** (the pipeline mode was sync-only, a problem for DB-lookup rules). FluentValidation's *other* recommended path is plain **manual** `ValidateAsync` — acceptable, but rejected here because it repeats the validate/return boilerplate in every action.
+  - .NET 10's native `AddValidation()` is **Minimal-API + DataAnnotations + synchronous only** — not a substitute for FluentValidation here.
+- Invoke a validator explicitly **only** for a rule that cannot run in the model pipeline (e.g. it needs a service result already fetched inside the action). Keep that the exception, not the norm.
+- DataAnnotations are acceptable on Options classes (paired with `.ValidateDataAnnotations()` per the Options section) and on simple non-FluentValidation property checks. Don't mix the two for the **same** DTO.
+
+## JSON serialization (property naming)
+
+- **Set the wire naming convention once, globally**, via `JsonSerializerOptions.PropertyNamingPolicy` — never by decorating every property. The convention is **lower camelCase** (`JsonNamingPolicy.CamelCase`) — the ASP.NET Core Web default and the idiomatic JS/TS-friendly shape. Configure it once in the composition root:
+  ```csharp
+  // Minimal-API / endpoint serialization
+  builder.Services.ConfigureHttpJsonOptions(o =>
+      o.SerializerOptions.PropertyNamingPolicy = JsonNamingPolicy.CamelCase);
+  // MVC controllers
+  builder.Services.AddControllers()
+      .AddJsonOptions(o => o.JsonSerializerOptions.PropertyNamingPolicy = JsonNamingPolicy.CamelCase);
+  ```
+  DTO members stay plain PascalCase C# (`ServerNow`, `DeviceId`) and serialize **and deserialize** as `serverNow`, `deviceId` automatically.
+- **Migration note (BREAKING — not behavior-preserving).** The contract historically shipped `snake_case` (`server_now`, `device_id`, …), consumed raw by the SPA (`web/`), the TS types, E2E/blackbox tests, `TestCommon` DTOs, seed fixtures, and `_docs/`. Flipping the policy to camelCase renames **every field on the wire**, so it is a breaking change tracked as **its own ticket** and must land **atomically** with the SPA + tests + fixtures + docs update (and an API version bump). Do **not** flip the policy — or strip the snake_case attributes — in isolation, and never inside a "behavior-preserving" refactor task.
+- **`[JsonPropertyName("...")]` is for overrides only — names the global policy cannot derive — never the default way to set casing.** It always wins over the policy, so reach for it ONLY when:
+  - the wire name is **irregular** vs. what the policy produces — e.g. acronym casing the CamelCase policy only lowercases the first char of (`IPAddress` → `iPAddress`, `DeviceID` → `deviceID`) when the contract wants `ipAddress`/`deviceId`, or an external contract demands an exact string we don't control;
+  - the wire name is **not a valid C# identifier** or otherwise inexpressible by any policy.
+- Decorating every property with `[JsonPropertyName("...")]` to emulate a global policy is a **code-review-fail signal**: it is noise, it drifts, and it silently shadows the policy. If a whole DTO's attributes merely restate what the policy would produce, delete them and rely on the policy.
+- Enum string values use a `JsonStringEnumConverter`; keep its naming policy consistent with the property policy.
+- Grounding: Microsoft's System.Text.Json docs recommend the global `PropertyNamingPolicy` for project-wide conventions and reserve `[JsonPropertyName]` for exact-string overrides (it takes highest precedence and overrides the policy).
+
+## Error handling
+
+> General principle (cross-language): see `coderule.mdc` → "Error handling". This section is the .NET realization (the three-tier model, central handler, opaque-500, and status mapping all originate there).
+
+This project uses a **business-exception model with one central handler** — *not* `Result<T,E>` and *not* per-method `try/catch`. Three failure tiers, three treatments:
+
+1. **Input validation** — handled by the **auto-validation action filter, never by throwing.** FluentValidation auto-validation (see Validation) short-circuits the request before the action runs and returns the `400` (slug-keyed `problem+json` via the custom result factory). Do **not** raise a `ValidationException` for request-shape validation.
+2. **Business-rule violations** (expected, part of the API contract: not-found, conflict, invariant violation, forbidden-by-rule) — the service **throws a `BusinessException` subtype**. Services express failure by throwing; they do **not** return error-wrapper values and do **not** catch their own business exceptions.
+3. **Unexpected failures** (bugs — NRE, invariant breaks; infrastructure — DB unreachable, network) — thrown by the framework/runtime and left to **propagate** to the central handler.
+
+### Business exception hierarchy
+
+- A single abstract base — `abstract class BusinessException : Exception` — carries the HTTP mapping data: an `int Status` and a stable `string Slug` (and optional extension members). Every expected, contract-level failure is a concrete subtype that fixes its own status; **there is no single blanket business status code**:
+  - not-found → `404`
+  - state conflict (duplicate key, concurrent edit, illegal state transition) → `409`
+  - well-formed request that violates a business invariant → `422`
+  - forbidden by a business rule (not auth-scheme denial) → `403`
+- The `Slug`/`Status`/title **must reuse the existing `FleetViewerProblems` slug catalog** (`Common/Problems/`) so the `application/problem+json` wire contract (`type` URI, `title`, `status`, any `code` extension) stays byte-identical to what blackbox tests pin. The catalog stays the single source of truth for the error contract; the exception types reference it.
+- Choose `422` vs `409` by meaning, never interchangeably: `422` = the request is well-formed but the business invariant rejects it; `409` = it conflicts with the resource's current state.
+
+### Central handler (catch in exactly one place)
+
+- Register **one** `IExceptionHandler` via `builder.Services.AddExceptionHandler<...>()` + `AddProblemDetails()` + `app.UseExceptionHandler()`. It maps:
+  - `BusinessException` → `ProblemDetails` built from its `Status` + `FleetViewerProblems.TypePrefix + Slug` (+ extensions). **Do NOT log these as errors** — they are expected 4xx contract outcomes; at most a `Debug`/`Information` line. Logging them at `Error` pollutes the error rate and pages on-call for normal client mistakes.
+  - **everything else (unexpected)** → `500` `ProblemDetails` with a **fixed, opaque production body** — `title: "Unexpected error"`, `detail: "An unexpected error occurred. Our team has been notified."` — and **log the full exception to Serilog at `Error`** (`logger.LogError(ex, ...)`) with the correlation id, so the log entry correlates to the client's response. The body must **never** carry the exception message, stack trace, or any internal detail (information-disclosure risk). In `Development` only, it is acceptable to surface `ex.Message`/stack in the body to aid debugging — gate that on `IHostEnvironment.IsDevelopment()`.
+- **No per-method `try/catch` for error mapping.** A handler/controller does not catch business exceptions to turn them into responses — that is the central handler's only job. Legitimate local `catch` blocks remain only for: converting a third-party/framework exception into a `BusinessException` at a boundary, honoring `OperationCanceledException`, or keeping a background loop alive (catch-log-continue). Never an empty/silent catch (see `coderule.mdc`).
+- **Do not throw on hot per-item paths** (e.g. ingest per-record processing): exceptions are for request-level outcomes, not inner loops — return/skip with a counted metric there.
+- API error responses are always `ProblemDetails` (RFC 7807) with a stable slug `type` when the failure is part of the contract.
+
+## HttpClient
+
+- **Never `new HttpClient()` per request** (sockets enter `TIME_WAIT` for ~240s; you exhaust the ephemeral port range under load).
+- **Never use a naive `static HttpClient`** either (handlers don't rotate, DNS changes are missed).
+- Register via `IHttpClientFactory` — typed or named clients:
+  ```csharp
+  builder.Services.AddHttpClient<MyApiClient>(c => c.BaseAddress = new Uri("https://api.example.com"));
+  ```
+- **Don't capture a typed `HttpClient` in a singleton.** Typed clients are Transient; capturing one in a singleton defeats handler rotation. Inject `IHttpClientFactory` into the singleton and call `CreateClient(name)` per operation, **or** configure `SocketsHttpHandler.PooledConnectionLifetime` so DNS refreshes at the socket level instead of the factory level.
+
+## Modern C# / nullable reference types
+
+- Enable nullable reference types (`<Nullable>enable</Nullable>`) on every new project.
+- **Don't paper over NRT warnings with `!`** (null-forgiving operator). Prefer:
+  - `required` members (C# 11) for properties the caller must initialize via object initializer.
+  - Constructor parameters for invariants established at construction.
+  - `[NotNullWhen(true)]` / `[NotNull]` / `[MaybeNull]` attributes for `Try*` patterns.
+- Use `ArgumentNullException.ThrowIfNull(x)` at the top of any public method taking a reference-type argument. NRTs are design-time only; library entry points still need runtime guards.
+
+## Static classes and static members
+
+> General principle (cross-language): see `coderule.mdc` → "Static members (functions / classes)". Below is the .NET realization plus framework-specific exemptions.
+
+Default to **instance classes behind an interface, registered in DI and constructor-injected.** That is what makes a unit testable (mockable), swappable, and free of hidden global state. `static` is the exception, not the default — reach for it only when the alternative below clearly applies.
+
+**No business logic in a static method — ever.** `static` is for *mechanics* (convert, parse, compute, compare), never for *decisions* (what the system should do, which rule applies, what happens next). Domain logic lives in a service.
+
+- **`static` is appropriate ONLY for:**
+  - **Pure, stateless, and SIMPLE functions** — output depends solely on the arguments; no I/O, no clock, no `Random`/`Guid.NewGuid`, no DB/file/network, no mutable shared state; **and** the body is short and obvious (math, encoding/decoding, parsing, formatting, a small predicate). Simplicity — not purity alone — is the bar: the moment a would-be helper carries domain decisions, branches across many cases, or is complex enough to deserve its own unit-test suite, it stops being a "helper." Make it an **instance service behind an interface** so it is injectable, mockable by its collaborators, and discoverable. A complicated *pure* function still belongs in a service.
+  - **Extension methods** over framework or domain types, when the body is pure and simple (e.g. claim/identity readers, enum⇄wire mappers).
+  - **Constants / well-known values** (a `static class` holding `const`s).
+  - **Static factory methods** on a type (private ctor + `public static Create(...)` returning a fully-formed instance) — an accepted construction pattern, distinct from a static *service*.
+- **Never use `static` for:**
+  - **Business / domain logic of any kind**, even if currently it looks "pure." Decisions belong in a tested, injectable service.
+  - A helper that touches I/O, configuration, time, randomness, or any external system — that is a *service*. Define an interface, make it an instance class, inject it. A static method that reaches a DB/clock/file cannot be mocked and forces brittle integration-style tests.
+  - **Mutable static fields of any kind.** Global mutable state is a thread-safety and test-isolation hazard. A cache or in-memory state store belongs in a DI **singleton behind an interface**, never a `static Dictionary`.
+  - Avoiding `new`/DI "ceremony." DI registration is one line and buys testability; saving it is never a reason to go static.
+- **Controllers are instance classes (constructor DI), not static.** A controller is `[ApiController] public sealed class XxxController(IXxxService svc) : ControllerBase { ... }` — dependencies are constructor-injected, actions are thin, and the type is never `static`. This is the standard for all new HTTP code (see "Solution layout & layering").
+- **Transitional exemption — legacy Minimal-API endpoint classes.** Existing `internal static class XxxEndpoints` exposing `MapXxxEndpoints(this RouteGroupBuilder group)` + `static` handler methods are the idiomatic *Minimal-API* pattern (no static state; deps are per-request method parameters; testable via `WebApplicationFactory`) and are **not** a static-class violation **while they exist**. Where the codebase has chosen controllers, migrate them and do **not** add new ones; until migrated, keep handler bodies thin with logic in injected services.
+- The static-OK rule also covers framework callback types that the runtime instantiates or invokes by convention — `AuthenticationHandler<TOptions>`, middleware `InvokeAsync`, `CookieAuthenticationEvents`, route predicates. They legitimately receive `HttpContext`/framework primitives and are not "static-class" or "HttpContext-discipline" violations.
+- **Library-mandated process-global statics are an accepted exception.** Some libraries are *designed* around a process-global, thread-safe static registry — e.g. a metrics library's `static readonly` counter/gauge collectors, or a `static` logger handle. Those `static readonly` fields are not the "mutable static state" this rule bans; do not force them behind a bespoke interface. A stateless utility over the system CSPRNG is likewise acceptable as `static` (folding it behind an interface for consistency with sibling generators is a fine choice, not a requirement).
+
+## Data access (EF Core)
+
+> General principle (cross-language): see `coderule.mdc` → "Data access" (single ORM path, justify raw SQL, prevent N+1). Below is the EF Core realization.
+
+- **Use the project ORM (EF Core for this repo) as the ONLY data-access path for application reads/writes.** Raw SQL via `CommandText`, `FromSqlRaw`, `FromSqlInterpolated`, `ExecuteSqlRaw`, `ExecuteSqlInterpolated`, or `NpgsqlCommand`/`NpgsqlConnection.CreateCommand()` is **forbidden by default** in endpoint, service, and repository code. Reaching for raw SQL because "it's simpler" or "EF generates ugly SQL" is not a valid reason — write the LINQ query, profile if you must, and only then justify a workaround.
+  - Narrow exceptions (each requires a 1-line comment in the code naming the EF limitation being worked around):
+    - **DDL the ORM cannot express** — `CREATE EXTENSION`, vendor enum-cast DEFAULT (`HasDefaultValueSql("'active'::device_state")`). Confine to migrations or to one-shot `IHostedService.StartAsync` bootstrap hooks.
+    - **Vendor-specific operators / functions** (e.g., TimescaleDB `time_bucket`, `make_interval(secs => ...)`, hypertable functions, PostGIS `ST_*`). Wrap each operator in a single repository method behind an interface; nowhere else in the codebase touches raw SQL for that operator. Prefer EF Core function mapping (`HasDbFunction` + `[DbFunction]`) before falling back to `FromSqlInterpolated`.
+    - **Benchmarked hot path** where EF demonstrably generates a worse plan than hand-rolled SQL. Requires a `BenchmarkDotNet` file checked in next to the workaround proving the gap. "We think it's faster" is not a benchmark.
+  - Prevent N+1 with `.Include()` / projection / explicit `.Select()`. New raw-SQL sites that do not fit one of the three exceptions MUST be flagged in code review as **High** severity (Maintainability / Architecture). Reviewers reject the PR until the SQL is either replaced with LINQ or moved behind a justified repository method with the required comment.
+- **`AsNoTracking()` on every read-only query.** The change tracker costs ~50% more memory and 2.9–5.2× more time on typical reads; you pay it for nothing on `GET` endpoints, reports, lookups. For read-heavy services, set `QueryTrackingBehavior.NoTracking` as the DbContext default and opt **in** to tracking with `.AsTracking()` on update paths.
+
+## ASP.NET Core handler discipline (controllers)
+
+> General principle (cross-language): see `coderule.mdc` → "Boundary discipline" (don't leak request/response context into business logic; authorize once at the boundary). Below is the ASP.NET Core realization.
+
+These rules keep controller actions and services free of framework primitives that hide dependencies, defeat unit testing, and bypass the auth/binding pipelines the framework already gives you. (They also apply to the legacy Minimal-API handlers still being migrated.)
+
+### `HttpContext` discipline
+
+- **Do not pass `HttpContext`, `HttpRequest`, `HttpResponse`, or `IHttpContextAccessor` into services or repositories.** Extract the values you need (headers, route values, body, `ClaimsPrincipal`) inside the handler and pass them down as typed parameters.
+- Take `HttpContext` (or `HttpRequest`/`HttpResponse`) as a handler parameter **only** when no binding source can express the requirement. Concrete examples that justify it:
+  - Custom body framing or streaming (you read `Request.Body`/`BodyReader` yourself).
+  - Multiple discriminated payload shapes on one URL that cannot be one DTO.
+  - Pre-allocation size caps that must reject **before** the body materializes into objects.
+  - Writing a custom response envelope that doesn't fit `Results.*`/`TypedResults.*`.
+  Document the reason with a `//` comment on the parameter or above the method.
+- Prefer **separate endpoints/methods** over discriminated payload shapes on one URL. Only fuse them when splitting would duplicate the majority of the validation logic — otherwise you trade testability for one fewer route registration, which is rarely worth it.
+- Default to specific binding sources: `[FromBody]`, `[FromQuery]`, `[FromHeader]`, `[FromRoute]`, `[FromServices]`, `ClaimsPrincipal user`, `CancellationToken cancellationToken`. Each of those is documented, testable, and integrates with OpenAPI.
+
+### JSON deserialization
+
+- **Default to `[FromBody]` + a typed `record`/DTO.** The framework calls `JsonSerializer.DeserializeAsync` for you, validates `Content-Type`, surfaces `BadHttpRequestException` on malformed input, and produces OpenAPI metadata.
+- Direct `JsonDocument` / `Utf8JsonReader` parsing of `Request.Body` is allowed **only** when typed deserialization cannot express the required validation. Allowed reasons:
+  - **Typed slug-keyed error envelopes** that the standard binder cannot produce (e.g., per-field problem+json with a stable `type` URI).
+  - **Pre-allocation size caps** that must reject `batch-too-large` before the array materializes.
+  - **Shape discrimination at parse time** when the alternative is a single fat DTO + runtime branching.
+  Each site needs a one-line comment naming which exception applies.
+- Reading raw `Request.Body` for plain typed JSON content is a code-review-fail signal in the absence of one of the named exceptions.
+
+### Custom authentication schemes
+
+- Custom bearer/token/API-key schemes go through **`AuthenticationHandler<TOptions>`** registered via `AddAuthentication().AddScheme<TOptions, THandler>(name, …)`. Apply `.RequireAuthorization(new AuthorizeAttribute { AuthenticationSchemes = name })` or `[Authorize(AuthenticationSchemes = name)]` on the endpoint.
+- **Do not read `Authorization` / cookie / API-key headers manually inside a handler that is `.AllowAnonymous()`.** That bypasses the auth pipeline, makes the auth logic unreusable for any second endpoint, and forces tests to reach the logic via reflection.
+- If you need a custom 401/403 body envelope (e.g. typed `application/problem+json` with a slug), override `HandleChallengeAsync` / `HandleForbiddenAsync` in the scheme handler — not by bypassing the pipeline.
+- In the endpoint, take `ClaimsPrincipal user` as a parameter and read identity from claims (`user.FindFirstValue(...)`). The auth handler is responsible for putting the right claims on the principal.
+
+### Authorization (declare-once at the boundary)
+
+- Authorize at the **boundary, once** — not per action. In MVC, put `[Authorize(Policy = "...")]` on the **controller class** (or a shared base controller); every action inherits it. Override on a single action with a narrower `[Authorize(Policy = ...)]` / `[AllowAnonymous]` only where it genuinely differs.
+- The Minimal-API equivalent is `group.MapGroup("/...").RequireAuthorization(policy)` on the **route group**. Both compile to the **same authorization metadata** — the group-level fluent call and the class-level attribute are equally correct and equally DRY. Per-method attributes / per-endpoint `RequireAuthorization` are for intentional per-route overrides only.
+- Name policies centrally (a single constants holder) and reference the constant — never inline role strings at the call site.
+
+### Current-user / identity access
+
+- **Inject `ClaimsPrincipal` directly into handlers for current-user identity; read it through the shared `ClaimsPrincipalExtensions` (`GetUserId()`, `GetSessionId()`, `GetDeviceId()`).** Do **not** wrap identity access in an `ICurrentUser` / `ICurrentUserProvider` service by default.
+- Why `ClaimsPrincipal` is the right seam here (not an over-coupling):
+  - It is a **data-driven seam whose producer is the auth handler** — the cookie scheme, `DeviceBearerAuthenticationHandler`, or any future JWT all populate the *same* `ClaimsPrincipal`. The handler is already decoupled from *how* identity was obtained.
+  - It is **available for free** in the HTTP layer — `ControllerBase.User` in a controller action (or a `ClaimsPrincipal user` parameter in a legacy Minimal-API handler), sourced from `HttpContext.User`; no `IHttpContextAccessor`, no scoped registration, no lifetime caveat. Identity stays in the `Api` layer: a controller reads `User`, extracts the IDs it needs via `ClaimsPrincipalExtensions`, and passes **plain values** (`Guid userId`) into the service — `ClaimsPrincipal` does not cross into the Services layer.
+  - It is **testable without an interface**: `ClaimsPrincipal` is `new`-able with arbitrary claims and its behaviour (`IsInRole`, `FindFirst`, the extensions) is fully driven by those claims. Construct a real principal with test claims — preferable to a mocked `IPrincipal`, which can diverge from real claim-matching semantics. (In this repo, handlers are exercised over HTTP via `WebApplicationFactory` with a real login, so identity is never substituted anyway.)
+  - The `ClaimsPrincipalExtensions` already provide the domain-friendly, centralized read surface that a provider's properties would duplicate.
+- A current-user provider adds a scoped `IHttpContextAccessor`-backed service — exactly the captive-dependency shape the DI section warns about — to replace a free, already-abstracted, already-testable binding. That fails the "simplicity is the highest priority" bar unless one of the concrete triggers below holds.
+- **Introduce an `ICurrentUser` abstraction ONLY when a named trigger appears:**
+  1. **Identity is needed outside an HTTP request** — background job, message consumer, worker thread — where `ClaimsPrincipal` cannot be bound from the pipeline. A provider with swappable impls (HTTP-backed vs job-context) earns its keep.
+  2. **The domain layer must consume identity** and you do not want `System.Security.Claims` types leaking into domain code — expose a domain-pure `ICurrentUser` value instead.
+  3. **You need richer-than-claims current-user data** (a loaded `User` entity, tenant, permission set) resolved and cached per request.
+  When introduced: back the HTTP implementation with `IHttpContextAccessor`, register it **Scoped**, never capture it in a singleton, and keep `ClaimsPrincipalExtensions` as the implementation detail it delegates to.
+
+### Response shapes
+
+**Controllers (the standard here): default to `ActionResult<T>`.** It mixes the success type `T` with `ActionResult` error shapes, participates in MVC's configured output formatters / content negotiation, and is the most reliable for OpenAPI:
+- Annotate with `[ProducesResponseType]`; the `Type` can be **omitted for the success code** (`[ProducesResponseType(StatusCodes.Status200OK)]`) — it is inferred from `T`. Add one attribute per additional status code (`404`, `409`, …).
+- Return the value directly (`return product;` — implicit cast to `200 OK`) or a `ControllerBase` helper for other shapes (`NotFound()`, `Conflict()`, `BadRequest(error)`, `CreatedAtAction(...)`).
+- The auto-validation action filter already produces the `400` for invalid input before the action runs (see Validation) — don't hand-write that path.
+- Keep the action **thin**: it maps the service's **success value** onto the success shape (`return product;` → `200`, `CreatedAtAction(...)` → `201`) and does not compute the business decision itself. **Expected failures are not mapped here** — the service throws a `BusinessException` subtype and the central `IExceptionHandler` produces the `ProblemDetails` (see Error handling). So a controller action has essentially no error branches: happy path in, success shape out.
+- `TypedResults` / `Results<T1, T2, …>` / `IResult` **are** usable in controllers, but they are the *Minimal-API* idiom and they **bypass MVC's configured output formatters / content negotiation** (they write the response directly — Microsoft Learn: "Does not leverage the configured Formatters"). Prefer `ActionResult<T>` in a controller; reach for `IResult` only for a deliberately format-agnostic raw response.
+
+**Legacy Minimal-API endpoints (until migrated): default to `TypedResults.*`** over `Results.*`. `TypedResults` returns concrete types (`Ok<T>`, `NotFound`, `BadRequest<T>`) that carry OpenAPI metadata and are unit-testable without casting. For handlers that return more than one shape, declare the return type as `Results<T1, T2, ...>` — the compiler enforces every branch returns a declared type and the OpenAPI generator reads the union, so no `Produces`/`ProducesResponseType` attributes are needed:
+  ```csharp
+  app.MapGet("/items/{id}", Results<Ok<Item>, NotFound> (int id) =>
+      item is not null ? TypedResults.Ok(item) : TypedResults.NotFound());
+  ```
+  Don't mix `Results.*` and `TypedResults.*` in the same handler — you lose the metadata.
+
+### Service results vs. wire envelopes
+
+> General principle (cross-language): see `coderule.mdc` → "Architecture & layering › Service results vs. transport envelopes". Below is the .NET realization.
+
+- A service returns a **domain result** — a record of the values it computed (`IReadOnlyList<LiveDevice>`, a small snapshot record) on success, and **throws a `BusinessException` subtype** on an expected failure (see Error handling); it does not return error-wrapper values. The **controller maps the success value onto the wire DTO**. The response envelope (the `*Response` record, its field names, the HTTP status) is an **Api-layer concern**; the domain result is not, and ASP.NET / wire types must not appear in a service signature (see "Solution layout & layering").
+- **A value that the response echoes to the client but that the service ALSO used to compute the result is owned by the service** — it returns that value alongside the data; the controller must NOT independently re-derive it. Two clocks/sources for the same conceptual value is a latent bug.
+  - Canonical case: a "server now" timestamp that a projection uses to decide freshness/staleness (which devices are dropped, what color each gets) **and** is echoed so the client renders relative ages consistently. If the controller stamped its own `DateTimeOffset.UtcNow`, it would diverge from the instant the service filtered against — a boundary bug.
+  - Pattern: the service injects `TimeProvider`, captures the instant **once**, uses it, and returns it inside a domain result — e.g. `LiveSnapshot(DateTimeOffset CapturedAt, IReadOnlyList<LiveDevice> Devices)`. The controller returns `ActionResult<LiveStateResponse>`, mapping `CapturedAt → server_now`. The envelope name and JSON shape stay in the Api layer; the *instant* originates in the Services layer where it is consumed.
+- The opposite case: a value that is **purely an HTTP/transport artifact and is never consumed by domain logic** (a `Location` header, a per-response correlation id minted for tracing) is owned by the **Api layer** and the service never sees it.
+- Heuristic: ask "does the business logic *read* this value to make a decision?" If yes → it lives in the service and is returned. If it is only *formatting/transport* → it lives in the controller.
+
+## Testing
+
+> General principle (cross-language): see `coderule.mdc` → "Testing (real dependencies)" (real engine over fakes for query-correctness; share expensive fixtures). Below is the .NET realization.
+
+- **xUnit** is the test framework for this repo. Use its per-test class lifecycle (constructor = setup, `IDisposable.Dispose` / `IAsyncLifetime.DisposeAsync` = teardown) — that's what most integration-testing patterns assume.
+- **FluentAssertions** for assertions: `result.Should().Be(...)`, `collection.Should().HaveCount(3).And.ContainSingle(x => ...)`, etc. Failure messages are much clearer than raw `Assert.Equal`, and the fluent chain reads like the spec it tests.
+- **`WebApplicationFactory<Program>`** for ASP.NET Core integration tests. It boots the real DI container and pipeline from `Program.cs` in-memory. Expose `Program` to the test project with `public partial class Program;` in `Program.cs`. Share the factory across tests in a class with `IClassFixture<T>` and across classes with `ICollectionFixture<T>` — host-boot is the expensive step; don't re-pay it per test.
+- **Never use the EF Core in-memory provider for query-correctness tests.** Its semantics diverge from real Postgres/SQL Server (LINQ translation differences, no real transactions, no concurrency tokens). Use Testcontainers (real Postgres container via `IAsyncLifetime` on the factory) + Respawn for between-test cleanup. The in-memory provider is acceptable only for fast smoke tests where you're not asserting query shape.
+- Tests follow the Arrange / Act / Assert pattern with `// Arrange` / `// Act` / `// Assert` comments (workspace convention; see `coderule.mdc`).
+
+## Cross-cutting
+
+- Use middleware for cross-cutting: auth, error handling, logging. Standard order in `Program.cs`: forwarded headers → exception handler → HTTPS/HSTS → static files → routing → CORS → authentication → authorization → rate limiter → endpoints.

.cursor/rules/large-file-writes.mdc

@@ -0,0 +1,41 @@
+---
+description: "Use chunked writes (Write + StrReplace marker pattern) for large generated files, especially after a monolithic Write fails"
+alwaysApply: true
+---
+# Large File Writes — Chunk on Failure
+
+When a `Write` call to a single file fails (timeout, payload limit, "Invalid arguments", or any tool error) and the intended content is large (>~500 lines or >~50 KB), do NOT retry the same monolithic Write. Switch to chunked writes:
+
+1. **First Write** — create the file with header + table of contents (if applicable) + an explicit append marker, e.g.
+
+   ```
+   <!-- INSERTION_POINT do-not-remove-until-final-chunk -->
+   ```
+
+2. **Each subsequent chunk** — use `StrReplace` to replace the marker with `<new content>\n<marker>` so the marker stays at the end. This is idempotent: if a chunk fails, retry it without losing earlier chunks.
+
+3. **Final chunk** — `StrReplace` removes the marker.
+
+## Why
+
+- Tool argument size limits and transient failures hit large monolithic writes hardest. Retrying the same large payload typically fails for the same reason.
+- Chunked writes are recoverable per chunk. The earlier chunks are durable on disk.
+- A unique marker is greppable, visible in diffs, and stops accidental insertion in the wrong place.
+
+## Triggers
+
+- Generated documentation that aggregates per-component content (epics, design docs, multi-section architecture summaries, traceability dumps).
+- Large fixture or test-data files written from a template.
+- Any single-file artifact you can pre-estimate at >~500 lines.
+
+## Do NOT chunk
+
+- Files under ~200 lines — a single `Write` is faster, clearer, and easier to review.
+- Source code files where appending breaks module structure (functions, classes, imports). Split into multiple files instead.
+- Files where ordering of sections is computed late and inserting in the middle is required — use a single `Write` once the full content is known.
+
+## Anti-patterns
+
+- Retrying the same failed monolithic `Write` more than once. Twice is the limit; on the second failure, switch strategies.
+- Using `Shell` with heredoc (`cat <<EOF`) or `echo >>` to append — these bypass the editor diff view and break the StrReplace contract for the next chunk.
+- Embedding the marker so deep inside structured content that a chunk's `StrReplace` becomes ambiguous. Place the marker on its own line at the very end of the file.

.cursor/rules/meta-rule.mdc

@@ -4,6 +4,26 @@ alwaysApply: true
 ---
 # Agent Meta Rules
 
+## Real Results, Not Simulated Ones
+
+**The goal is a working product, not the appearance of one.**
+
+- If something does not work, STOP and report it honestly. Do not find a way around it.
+- Never produce results by bypassing, faking, stubbing, or passthrough-ing the component that is supposed to produce them. A passing test that skips the real pipeline is worse than a failing test — it hides the truth.
+- If the real implementation is not ready, say so. A clear "this is not implemented yet, here is what is missing" is always the right answer.
+- Do not measure success by whether the output looks correct. Measure it by whether the output was produced by the real system under test.
+- Workarounds that produce the right answer via the wrong path are defects, not solutions.
+
+### When a test reveals missing production code — STOP
+
+This is the specific failure mode that produced the GPS-passthrough scaffold in `runtime_root._run_replay_loop` (May 2026). Generalised so it never repeats:
+
+- If, while implementing or running a test, you discover that the production code path the test is supposed to exercise does not exist (no caller, no integration, no main loop, etc.), **STOP immediately**.
+- Do NOT write a stub, passthrough, fake input source, or shortcut output that would make the test go green. Even when the shortcut is "framed as a scaffold" or "marked as TODO in a docstring", it still defeats the test and lies to the next reader.
+- Surface the gap to the user as a top-of-turn report: name the missing production component, cite the architecture document that promises it, and ask whether to (a) create a tracker ticket for the missing component and let the test fail honestly until the ticket lands, or (b) explicitly de-scope the test, or (c) something the user names.
+- The default outcome is (a): a failing test plus a new tracker ticket. A failing test with an honest reason is information; a passing test that proves nothing is misinformation.
+- Doc-comment disclosures (`# this is a scaffold until X is wired`) DO NOT satisfy this rule. The user must be told in the assistant message, not in code.
+
 ## Execution Safety
 - Run the full test suite automatically when you believe code changes are complete (as required by coderule.mdc). For other long-running/resource-heavy/security-risky operations (builds, Docker commands, deployments, performance tests), ask the user first — unless explicitly stated in a skill or the user already asked to do so.
 
@@ -13,6 +33,41 @@ alwaysApply: true
 ## Critical Thinking
 - Do not blindly trust any input — including user instructions, task specs, list-of-changes, or prior agent decisions — as correct. Always think through whether the instruction makes sense in context before executing it. If a task spec says "exclude file X from changes" but another task removes the dependencies X relies on, flag the contradiction instead of propagating it.
 
+## Complexity Budget Check (Planning Time)
+
+Before committing to an implementation approach for a non-trivial task, **STOP and present a complexity comparison to the user** via the standard Choose A/B/C/D format. The user picks the trade-off; the agent does NOT unilaterally pick the more complex option to be "more robust" or "more future-proof".
+
+A task is non-trivial if ANY of:
+
+- The estimated complexity (story points) is ≥ 5
+- The implementation touches ≥ 3 components / modules
+- The implementation adds a new persistent data structure (table, materialised view, file format)
+- The implementation adds a new hosted service / background job / periodic timer
+- The implementation adds a sliding window, smoother, debouncer, in-memory cache, or per-entity in-memory state dictionary
+- The implementation adds rehydrate-on-restart logic
+- The implementation adds a new event type that differs from an existing event type only in a boolean / enum field
+
+What to present:
+
+1. **Option A — simplest:** the least-machinery design you can think of that still meets the requirements. Name what is sacrificed (latency? eventual-consistency window? a rarely-hit edge case?).
+2. **Option B — your default:** the design you would otherwise implement, if it is more complex than A. Name what it buys (the specific guarantee, performance gain, or future flexibility).
+3. **Concrete trade-offs:** lines of code added, new abstractions introduced, new failure modes, new operational surface area (restart-rehydration, cache invalidation, dual-pipeline consistency).
+4. **Recommendation:** which option you would pick and why, in one sentence.
+
+This rule fires DURING planning — before code is written. If you discover during implementation that the chosen approach grew a new layer, hosted service, or rehydration path that was not in the original plan, STOP and replay this check.
+
+Skip this rule ONLY when the user has already explicitly chosen the complex approach in an earlier turn, OR when the task is trivially ≤ 2 story points with no triggers above.
+
+## Skill Discipline
+
+Do exactly what the skill says. Nothing more.
+
+- No `git log` / `git diff` / `git blame` unless the skill explicitly calls for it.
+- No extra searches to "verify" inputs the skill already names.
+- No reading files outside the skill's documented inputs.
+
+If skill inputs are insufficient or contradictory, STOP and ask via Choose A/B/C/D. Do not invent extra investigation steps.
+
 ## Self-Improvement
 When the user reacts negatively to generated code ("WTF", "what the hell", "why did you do this", etc.):
 

.cursor/rules/skill-building.mdc

@@ -0,0 +1,38 @@
+---
+description: "Standards for creating and maintaining Cursor skills"
+globs: [".cursor/skills/**"]
+---
+
+# Skill Building
+
+## When To Create A Skill
+- Create a skill for repeatable, bounded workflows that benefit from a reusable process.
+- Do not create a skill for a one-off task, vague goal, or workflow that still needs product decisions.
+- Start small; evolve the skill when repeated use reveals clearer steps, constraints, or checks.
+
+## Skill Contract
+- `SKILL.md` must define a clear `name` and a proactive `description` that explains when the skill should be used.
+- State expected inputs, constraints, workflow steps, and final output shape.
+- Make trigger conditions explicit enough that the agent can recognize intent without an exact command.
+- Base instructions on observable project evidence; do not invite fabrication or unsupported assumptions.
+
+## Keep The Core Lean
+- Keep `SKILL.md` concise and under the repo's `.cursor/` size guidance.
+- Move detailed standards, examples, and background knowledge into `references/`.
+- Put reusable output shapes in `templates/` or other skill-local assets instead of embedding them in the main instructions.
+- Keep one primary responsibility per skill; use an orchestrator skill only when multiple existing skills must run in a defined order.
+
+## Deterministic Work
+- Use scripts for mechanical steps that are repeatable, parameterized, and safer outside the model's reasoning.
+- Scripts must expose explicit inputs, avoid hidden side effects, and fail loudly on errors.
+- Do not use scripts to bypass review, hide destructive behavior, or hardcode secrets.
+
+## Quality Proof
+- Include realistic examples, checklists, or eval-style scenarios that define what good output looks like.
+- Cover common failure cases such as missing sections, leftover placeholders, hallucinated facts, unsafe actions, or malformed output.
+- Review skill changes against those checks before treating the skill as ready.
+
+## Security Review
+- Treat third-party skills like untrusted code until reviewed.
+- Inspect scripts, dependencies, references, secret handling, network calls, and destructive commands before use.
+- Prefer local, project-scoped assets and dependencies; document any external dependency the skill requires.

.cursor/rules/testing.mdc

@@ -8,8 +8,16 @@ globs: ["**/*test*", "**/*spec*", "**/*Test*", "**/tests/**", "**/test/**"]
 - One assertion per test when practical; name tests descriptively: `MethodName_Scenario_ExpectedResult`
 - Test boundary conditions, error paths, and happy paths
 - Use mocks only for external dependencies; prefer real implementations for internal code
-- Aim for 75%+ coverage on business logic; 100% on critical paths (code paths where a bug would cause data loss, security breaches, financial errors, or system outages — identify from acceptance criteria marked as must-have or from security_approach.md). The 75% threshold is canonical — see `cursor-meta.mdc` Quality Thresholds.
+- Aim for 75%+ coverage on business logic; **90% floor / 100% aim on critical paths** (code paths where a bug would cause data loss, security breaches, financial errors, or system outages — identify from acceptance criteria marked as must-have or from `security_approach.md`). 90% is the enforcement floor (blocking in CI / refactor verification / release pre-flight); 100% is the aspirational aim — drift below 100% but at-or-above 90% is acceptable. Both numbers are canonical — see `cursor-meta.mdc` Quality Thresholds.
 - Integration tests use real database (Postgres testcontainers or dedicated test DB)
 - Never use Thread Sleep or fixed delays in tests; use polling or async waits
 - Keep test data factories/builders for reusable test setup
 - Tests must be independent: no shared mutable state between tests
+
+## Test environment (this project)
+
+- **Unit tests** (`tests/unit/`): may run locally on the dev workstation (`pytest tests/unit/` in the project venv). Local PASS is equivalent to Jetson PASS for this tier because the suite is fully synthetic.
+- **Blackbox / e2e / performance / resilience / security / resource-limit** tests (`tests/e2e/`, `e2e/tests/`, `tests/perf/`, …): MUST run on the Jetson Orin Nano Super (or a Jetson-equivalent arm64 agent). Use `scripts/run-tests-jetson.sh` for local dev; CI runs `.woodpecker/01-test.yml` on the colocated arm64 Jetson Woodpecker agent.
+- Do NOT run e2e tests on the local workstation and report the result. If the Jetson is unreachable, the e2e verdict is "not run" — record the gap in `_docs/_process_leftovers/` rather than substituting a local result.
+- Tests gated by `RUN_REPLAY_E2E` or `@pytest.mark.tier2` are expected to SKIP locally; that is correct behaviour, not a failure to investigate.
+- Canonical source for this policy: `_docs/02_document/tests/environment.md` § Where each tier runs (active policy).

.cursor/rules/tracker.mdc

@@ -14,11 +14,14 @@ alwaysApply: true
 - Issue types: Epic, Story, Task, Bug, Subtask
 
 ## Tracker Availability Gate
-- If Jira MCP returns **Unauthorized**, **errored**, **connection refused**, or any non-success response: **STOP** tracker operations and notify the user via the Choose A/B/C/D format documented in `.cursor/skills/autodev/protocols.md`.
+- If Jira MCP returns **Unauthorized**, **errored**, **connection refused**, **timeout**, a non-2xx status code, an empty body, or any response shape that does not clearly confirm the requested change: **STOP IMMEDIATELY** — no automatic retry, no silent continuation. Surface the full raw error/response to the user verbatim and notify via the Choose A/B/C/D format documented in `.cursor/skills/autodev/protocols.md`.
+- A minimal `{"success": true}` body with no echoed issue state is NOT a confirmed transition. When a transition's success matters (status moves, ticket creation, blocking link), follow it with a read-back call (`getJiraIssue` or equivalent) and confirm the new state matches what you asked for. If the read-back disagrees → STOP and ASK.
+- Do NOT loop "retry up to N times before asking". One call, one verification. On failure, the user decides whether to retry.
 - The user may choose to:
-  - **Retry authentication** — preferred; the tracker remains the source of truth.
+  - **Retry the same operation** — once, after the user authorizes it. If it fails again, surface both responses.
+  - **Retry authentication** — preferred when the failure looks like an auth/credentials problem; the tracker remains the source of truth.
   - **Continue in `tracker: local` mode** — only when the user explicitly accepts this option. In that mode all tasks keep numeric prefixes and a `Tracker: pending` marker is written into each task header. The state file records `tracker: local`. The mode is NOT silent — the user has been asked and has acknowledged the trade-off.
-- Do NOT auto-fall-back to `tracker: local` without a user decision. Do not pretend a write succeeded. If the user is unreachable (e.g., non-interactive run), stop and wait.
+- Do NOT auto-fall-back to `tracker: local` without a user decision. Do not pretend a write succeeded. Do not paper over an opaque response by moving on. If the user is unreachable (e.g., non-interactive run), stop and wait.
 - When the tracker becomes available again, any `Tracker: pending` tasks should be synced — this is done at the start of the next `/autodev` invocation via the Leftovers Mechanism below.
 
 ## Leftovers Mechanism (non-user-input blockers only)

.cursor/skills/autodev/SKILL.md

@@ -1,9 +1,9 @@
 ---
 name: autodev
 description: |
-  Auto-chaining orchestrator that drives the full BUILD-SHIP workflow from problem gathering through deployment.
+  Auto-chaining orchestrator that drives the full BUILD → SHIP → EVOLVE workflow from problem gathering through release and retrospective.
   Detects current project state from _docs/ folder, resumes from where it left off, and flows through
-  problem → research → plan → decompose → implement → deploy without manual skill invocation.
+  problem → research → plan (incl. ADRs) → test specs → decompose → implement → tests → docs sync → deploy → release → retrospective without manual skill invocation.
   Maximizes work per conversation by auto-transitioning between skills.
   Trigger phrases:
   - "autodev", "auto", "start", "continue"
@@ -15,7 +15,7 @@ disable-model-invocation: true
 
 # Autodev Orchestrator
 
-Auto-chaining execution engine that drives the full BUILD → SHIP workflow. Detects project state from `_docs/`, resumes from where work stopped, and flows through skills automatically. The user invokes `/autodev` once — the engine handles sequencing, transitions, and re-entry.
+Auto-chaining execution engine that drives the full BUILD → SHIP → EVOLVE workflow. Detects project state from `_docs/`, resumes from where work stopped, and flows through skills automatically. The user invokes `/autodev` once — the engine handles sequencing, transitions, and re-entry.
 
 ## File Index
 
@@ -67,8 +67,9 @@ B3. Read state              — `_docs/_autodev_state.md` (if it exists).
 B4. Read File Index         — `state.md`, `protocols.md`, and the active flow file.
 
 ### Resolve (once per invocation, after Bootstrap)
-R1. Reconcile state         — verify state file against `_docs/` contents; on disagreement, trust the folders
-                               and update the state file (rules: `state.md` → "State File Rules" #4).
+R1. Reconcile state         — verify state file against `_docs/` contents; probe `<workspace-root>/../docs`
+                               (parent suite `docs/` — see `state.md` → "State File Rules" #4); on disagreement,
+                               trust the folders and update the state file (rules: `state.md` → "State File Rules" #4).
                                After this step, `state.step` / `state.status` are authoritative.
 R2. Resolve flow            — see §Flow Resolution above.
 R3. Resolve current step    — when a state file exists, `state.step` drives detection.
@@ -112,6 +113,15 @@ Do NOT modify, skip, or abbreviate any part of the sub-skill's workflow. The aut
 
 The state file (`_docs/_autodev_state.md`) is a minimal pointer — only the current step. See `state.md` for the authoritative template, field semantics, update rules, and worked examples. Do not restate the schema here — `state.md` is the single source of truth.
 
+**Conciseness rule (authoritative).** The state file MUST stay short. Acceptable content per field:
+
+- `name` — the step title from the active flow's Step Reference Table. That's it.
+- `sub_step.name` — kebab-case identifier from the active sub-skill. That's it.
+- `sub_step.detail` — **leave empty (`""`) by default.** Add a one-line note ONLY when the next-session resumer cannot infer where to pick up from `phase` + `name` + on-disk artifacts alone (e.g. `"batch 2 of 4"`, `"blocked on D-PROJ-2 reply"`, `"variant 1b"`). NEVER use `detail` as a changelog, recap, or summary of completed work — those facts belong in the relevant `_docs/` artifact (glossary, traceability matrix, leftovers folder, retro report, etc.) and in git history.
+- **Total file size target: <30 lines.** If you're tempted to write more, you're using the wrong artifact — write in `_docs/` instead.
+
+Multi-line `detail` blobs that recap what was just completed are a smell. The state file is a *pointer*, not a logbook.
+
 ## Trigger Conditions
 
 This skill activates when the user wants to:

.cursor/skills/autodev/flows/existing-code.md

@@ -3,7 +3,7 @@
 Workflow for projects with an existing codebase. Structurally it has **two phases**:
 
 - **Phase A — One-time baseline setup (Steps 1–8)**: runs exactly once per codebase. Documents the code, produces test specs, makes the code testable, writes and runs the initial test suite, optionally refactors with that safety net.
-- **Phase B — Feature cycle (Steps 9–17, loops)**: runs once per new feature. After Step 17 (Retrospective), the flow loops back to Step 9 (New Task) with `state.cycle` incremented.
+- **Phase B — Feature cycle (Steps 9–17, loops)**: runs once per new feature. After Step 17 (Retrospective), the flow loops back to Step 9 (New Task) with `state.cycle` incremented. Step 16.5 (Release) sits between Deploy (16) and Retrospective (17).
 
 A first-time run executes Phase A then Phase B; every subsequent invocation re-enters Phase B.
 
@@ -33,7 +33,8 @@ A first-time run executes Phase A then Phase B; every subsequent invocation re-e
 | 13 | Update Docs | document/SKILL.md (task mode) | Task Steps 0–5 |
 | 14 | Security Audit | security/SKILL.md | Phase 1–5 (optional) |
 | 15 | Performance Test | test-run/SKILL.md (perf mode) | Steps 1–5 (optional) |
-| 16 | Deploy | deploy/SKILL.md | Step 1–7 |
+| 16 | Deploy | deploy/SKILL.md | Step 1–7 (optional) |
+| 16.5 | Release | release/SKILL.md | Phase 1–6 (optional — only if Step 16 completed) |
 | 17 | Retrospective | retrospective/SKILL.md (cycle-end mode) | Steps 1–4 |
 
 After Step 17, the feature cycle completes and the flow loops back to Step 9 with `state.cycle + 1` — see "Re-Entry After Completion" below.
@@ -152,15 +153,17 @@ If `_docs/02_tasks/` subfolders have some task files already (e.g., refactoring
 ---
 
 **Step 6 — Implement Tests**
-Condition (folder fallback): `_docs/02_tasks/todo/` contains task files AND `_dependencies_table.md` exists AND `_docs/03_implementation/implementation_report_tests.md` does not exist.
+Condition (folder fallback): `_docs/02_tasks/todo/` contains test task files AND `_dependencies_table.md` exists AND `_docs/03_implementation/implementation_report_tests.md` does not exist.
 State-driven: reached by auto-chain from Step 5.
 
-Action: Read and execute `.cursor/skills/implement/SKILL.md`
+Action: Invoke `.cursor/skills/implement/SKILL.md` with task selection context **Test implementation**.
 
-The implement skill reads test tasks from `_docs/02_tasks/todo/` and implements them.
+The implement skill reads only test tasks from `_docs/02_tasks/todo/` and implements them.
 
 If `_docs/03_implementation/` has batch reports, the implement skill detects completed tasks and continues.
 
+For folder fallback, **test task files** means `*_test_infrastructure.md` plus task specs whose `**Component**` or `**Epic**` identifies `Blackbox Tests`.
+
 ---
 
 **Step 7 — Run Tests**
@@ -273,33 +276,63 @@ State-driven: reached by auto-chain from Step 14 (completed or skipped).
 Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Gate") with:
 - question:        `Run performance/load tests before deploy?`
 - option-a-label:  `Run performance tests (recommended for latency-sensitive or high-load systems)`
-- option-b-label:  `Skip — proceed directly to deploy`
+- option-b-label:  `Skip — proceed to deploy choice`
 - recommendation:  `A or B — base on whether acceptance criteria include latency, throughput, or load requirements`
 - target-skill:    `.cursor/skills/test-run/SKILL.md` in **perf mode** (the skill handles runner detection, threshold comparison, and its own A/B/C gate on threshold failures)
 - next-step:       Step 16 (Deploy)
 
 ---
 
-**Step 16 — Deploy**
+**Step 16 — Deploy (optional)**
 State-driven: reached by auto-chain from Step 15 (completed or skipped).
 
-Action: Read and execute `.cursor/skills/deploy/SKILL.md`.
+Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Gate") with:
+- question:        `Run deploy planning or refresh deploy artifacts for this cycle?`
+- option-a-label:  `Run deploy — update scripts/procedures for this release`
+- option-b-label:  `Skip — keep developing; deploy when ready for production`
+- recommendation:  `B during active feature work; A when this cycle should ship`
+- target-skill:    `.cursor/skills/deploy/SKILL.md`
+- next-step:       Step 16.5 (Release) — only when Step 16 was completed; otherwise Step 17 (Retrospective)
 
-After the deploy skill completes successfully, mark Step 16 as `completed` and auto-chain to Step 17 (Retrospective).
+On **skip**: mark Step 16 and Step 16.5 as `skipped`; auto-chain to Step 17 (Retrospective in cycle-end mode).
+
+On **complete**: mark Step 16 `completed` and auto-chain to Step 16.5 (Release).
+
+---
+
+**Step 16.5 — Release (optional)**
+State-driven: reached by auto-chain from Step 16 **only when Step 16 status is `completed`**, for the current `state.cycle`. If Step 16 was `skipped`, Step 16.5 is `skipped` and `/release` is not invoked.
+
+Action: Read and execute `.cursor/skills/release/SKILL.md`. The release skill owns its own user interaction (Phase 1 pre-release gate, Phase 2 strategy select, Phase 6 escalation). Autodev does NOT add a wrapping A/B/C gate. Pass cycle context (`cycle: state.cycle`).
+
+After the release skill exits, route on the verdict:
+
+- **Verdict `Released`** → mark Step 16.5 `completed` and auto-chain to Step 17 (Retrospective in cycle-end mode).
+- **Verdict `Released-with-override`** → mark Step 16.5 `completed` AND auto-chain to Step 17 (Retrospective in **incident mode**).
+- **Verdict `Rolled-Back`** → mark Step 16.5 `failed`. Auto-chain to Step 17 (Retrospective in **incident mode**). The cycle does NOT loop back to Step 9.
+- **Verdict `Aborted`** → mark Step 16.5 `not_started` (no live-system change) OR `failed` (live-system touched before abort). Surface the abort reason and STOP. Next `/autodev` invocation re-evaluates Phase B from the failed step.
 
 ---
 
 **Step 17 — Retrospective**
-State-driven: reached by auto-chain from Step 16, for the current `state.cycle`.
+State-driven: reached by auto-chain from Step 16.5 (any verdict) OR from Step 16/16.5 both `skipped`, for the current `state.cycle`.
 
-Action: Read and execute `.cursor/skills/retrospective/SKILL.md` in **cycle-end mode**. Pass cycle context (`cycle: state.cycle`) so the retro report and LESSONS.md entries record which feature cycle they came from.
+Action: Read and execute `.cursor/skills/retrospective/SKILL.md`. Mode selection:
 
-After retrospective completes, mark Step 17 as `completed` and enter "Re-Entry After Completion" evaluation.
+- Step 16.5 verdict `Released` → cycle-end mode
+- Step 16.5 verdict `Released-with-override` or `Rolled-Back` → incident mode
+
+Pass cycle context (`cycle: state.cycle`) so the retro report and LESSONS.md entries record which feature cycle they came from.
+
+After retrospective completes:
+
+- If Step 16.5 verdict was `Released` or `Released-with-override`, OR Step 16.5 was `skipped` → mark Step 17 as `completed` and enter "Re-Entry After Completion" evaluation (loop back to Step 9 for cycle N+1).
+- If Step 16.5 verdict was `Rolled-Back` → mark Step 17 as `completed` but do NOT loop back. Surface the incident retro path and STOP.
 
 ---
 
 **Re-Entry After Completion**
-State-driven: `state.step == done` OR Step 17 (Retrospective) is completed for `state.cycle`.
+State-driven: `state.step == done` OR Step 17 (Retrospective) is completed for `state.cycle` AND (Step 16.5 verdict was `Released` or `Released-with-override` OR Step 16.5 was `skipped`). A `Rolled-Back` cycle does NOT trigger Re-Entry — the user must explicitly invoke `/autodev` again.
 
 Action: The project completed a full cycle. Print the status banner and automatically loop back to New Task — do NOT ask the user for confirmation:
 
@@ -314,7 +347,7 @@ Action: The project completed a full cycle. Print the status banner and automati
 
 Set `step: 9`, `status: not_started`, and **increment `cycle`** (`cycle: state.cycle + 1`) in the state file, then auto-chain to Step 9 (New Task). Reset `sub_step` to `phase: 0, name: awaiting-invocation, detail: ""` and `retry_count: 0`.
 
-Note: the loop (Steps 9 → 17 → 9) ensures every feature cycle includes: New Task → Implement → Run Tests → Test-Spec Sync → Update Docs → Security → Performance → Deploy → Retrospective.
+Note: the loop (Steps 9 → 17 → 9) covers: New Task → Implement → Run Tests → Test-Spec Sync → Update Docs → Security → Performance → Deploy (optional) → Release (optional) → Retrospective. The cycle completes (and loops back to Step 9) on a `Released` or `Released-with-override` verdict, or when deploy/release were skipped; rolled-back or aborted releases stop the cycle.
 
 ## Auto-Chain Rules
 
@@ -341,9 +374,15 @@ Note: the loop (Steps 9 → 17 → 9) ensures every feature cycle includes: New
 | Test-Spec Sync (12, done or skipped) | Auto-chain → Update Docs (13) |
 | Update Docs (13) | Auto-chain → Security Audit choice (14) |
 | Security Audit (14, done or skipped) | Auto-chain → Performance Test choice (15) |
-| Performance Test (15, done or skipped) | Auto-chain → Deploy (16) |
-| Deploy (16) | Auto-chain → Retrospective (17) |
-| Retrospective (17) | **Cycle complete** — loop back to New Task (9) with incremented cycle counter |
+| Performance Test (15, done or skipped) | Auto-chain → Deploy choice (16) |
+| Deploy (16, completed) | Auto-chain → Release (16.5) |
+| Deploy (16, skipped) | Mark 16.5 `skipped` → auto-chain → Retrospective (17, cycle-end mode) |
+| Release (16.5, verdict Released) | Auto-chain → Retrospective (17, cycle-end mode) |
+| Release (16.5, verdict Released-with-override) | Auto-chain → Retrospective (17, **incident mode**) |
+| Release (16.5, verdict Rolled-Back) | Auto-chain → Retrospective (17, **incident mode**); cycle does NOT loop back |
+| Release (16.5, verdict Aborted) | STOP — surface abort reason; do not auto-chain |
+| Retrospective (17, after Released / Released-with-override / deploy skipped) | **Cycle complete** — loop back to New Task (9) with incremented cycle counter |
+| Retrospective (17, after Rolled-Back) | Cycle remains incomplete — STOP and surface incident retro path |
 
 ## Status Summary — Step List
 
@@ -379,9 +418,10 @@ Flow-specific slot values:
 | 14 | Security Audit             | — |
 | 15 | Performance Test           | — |
 | 16 | Deploy                     | — |
+| 16.5 | Release                  | `DONE (Released | Released-with-override | Rolled-Back | Aborted)` |
 | 17 | Retrospective              | — |
 
-All rows accept the shared state tokens (`DONE`, `IN PROGRESS`, `NOT STARTED`, `FAILED (retry N/3)`); rows 2, 4, 8, 12, 13, 14, 15 additionally accept `SKIPPED`.
+All rows accept the shared state tokens (`DONE`, `IN PROGRESS`, `NOT STARTED`, `FAILED (retry N/3)`); rows 2, 4, 8, 12, 13, 14, 15, 16, 16.5 additionally accept `SKIPPED`.
 
 Row rendering format (renders with a phase separator between Step 8 and Step 9):
 
@@ -404,5 +444,6 @@ Row rendering format (renders with a phase separator between Step 8 and Step 9):
  Step 14   Security Audit           [<state token>]
  Step 15   Performance Test         [<state token>]
  Step 16   Deploy                   [<state token>]
+ Step 16.5 Release                  [<state token>]
  Step 17   Retrospective            [<state token>]
 ```

.cursor/skills/autodev/flows/greenfield.md

@@ -1,6 +1,6 @@
 # Greenfield Workflow
 
-Workflow for new projects built from scratch. Flows linearly: Problem → Research → Plan → UI Design (if applicable) → Decompose → Implement → Run Tests → Security Audit (optional) → Performance Test (optional) → Deploy → Retrospective.
+Workflow for new projects built from scratch. Flows linearly: Problem → Research → Plan → UI Design (if applicable) → Test Spec → Decompose → Implement + Product Completeness Gate → Code Testability Revision → Decompose Tests → Implement Tests → Run Tests → Test-Spec Sync → Update Docs → Security Audit (optional) → Performance Test (optional) → Deploy (optional) → Release (optional, only if Deploy ran) → Retrospective.
 
 ## Step Reference Table
 
@@ -8,15 +8,22 @@ Workflow for new projects built from scratch. Flows linearly: Problem → Resear
 |------|------|-----------|-------------------|
 | 1 | Problem | problem/SKILL.md | Phase 1–4 |
 | 2 | Research | research/SKILL.md | Mode A: Phase 1–4 · Mode B: Step 0–8 |
-| 3 | Plan | plan/SKILL.md | Step 1–6 + Final |
+| 3 | Plan | plan/SKILL.md | Step 1, 2, 3, 4, 4.5 (ADR Capture), 5, 6 + Final |
 | 4 | UI Design | ui-design/SKILL.md | Phase 0–8 (conditional — UI projects only) |
-| 5 | Decompose | decompose/SKILL.md | Step 1–4 |
-| 6 | Implement | implement/SKILL.md | (batch-driven, no fixed sub-steps) |
-| 7 | Run Tests | test-run/SKILL.md | Steps 1–4 |
-| 8 | Security Audit | security/SKILL.md | Phase 1–5 (optional) |
-| 9 | Performance Test | test-run/SKILL.md (perf mode) | Steps 1–5 (optional) |
-| 10 | Deploy | deploy/SKILL.md | Step 1–7 |
-| 11 | Retrospective | retrospective/SKILL.md (cycle-end mode) | Steps 1–4 |
+| 5 | Test Spec | test-spec/SKILL.md | Phases 1–4 |
+| 6 | Decompose | decompose/SKILL.md (implementation task decomposition) | Step 1 + Step 1.5 + Step 2 + Step 4 |
+| 7 | Implement | implement/SKILL.md | Batch loop + Product Implementation Completeness Gate |
+| 8 | Code Testability Revision | refactor/SKILL.md (guided mode) | Phases 0–7 (conditional) |
+| 9 | Decompose Tests | decompose/SKILL.md (tests-only) | Step 1t + Step 3 + Step 4 |
+| 10 | Implement Tests | implement/SKILL.md | (batch-driven, no fixed sub-steps) |
+| 11 | Run Tests | test-run/SKILL.md | Steps 1–4 |
+| 12 | Test-Spec Sync | test-spec/SKILL.md (cycle-update mode) | Phase 2 + Phase 3 (scoped) |
+| 13 | Update Docs | document/SKILL.md (task mode) | Task Steps 0–5 |
+| 14 | Security Audit | security/SKILL.md | Phase 1–5 (optional) |
+| 15 | Performance Test | test-run/SKILL.md (perf mode) | Steps 1–5 (optional) |
+| 16 | Deploy | deploy/SKILL.md | Step 1–7 (optional) |
+| 16.5 | Release | release/SKILL.md | Phase 1–6 (optional — only if Step 16 completed) |
+| 17 | Retrospective | retrospective/SKILL.md (cycle-end mode) | Steps 1–4 |
 
 ## Detection Rules
 
@@ -80,12 +87,12 @@ If `_docs/02_document/` exists but is incomplete (has some artifacts but no `FIN
 ---
 
 **Step 4 — UI Design (conditional)**
-Condition (folder fallback): `_docs/02_document/architecture.md` exists AND `_docs/02_tasks/todo/` does not exist or has no task files.
+Condition (folder fallback): `_docs/02_document/architecture.md` exists AND `_docs/02_document/tests/traceability-matrix.md` does not exist.
 State-driven: reached by auto-chain from Step 3.
 
 Action: Read and execute `.cursor/skills/ui-design/SKILL.md`. The skill runs its own **Applicability Check**, which handles UI project detection and the user's A/B choice. It returns one of:
 
-- `outcome: completed` → mark Step 4 as `completed`, auto-chain to Step 5 (Decompose).
+- `outcome: completed` → mark Step 4 as `completed`, auto-chain to Step 5 (Test Spec).
 - `outcome: skipped, reason: not-a-ui-project` → mark Step 4 as `skipped`, auto-chain to Step 5.
 - `outcome: skipped, reason: user-declined` → mark Step 4 as `skipped`, auto-chain to Step 5.
 
@@ -93,34 +100,162 @@ The autodev no longer inlines UI detection heuristics — they live in `ui-desig
 
 ---
 
-**Step 5 — Decompose**
-Condition: `_docs/02_document/` contains `architecture.md` AND `_docs/02_document/components/` has at least one component AND `_docs/02_tasks/todo/` does not exist or has no task files
+**Step 5 — Test Spec**
+Condition (folder fallback): `_docs/02_document/FINAL_report.md` exists AND `_docs/02_document/architecture.md` exists AND `_docs/02_document/tests/traceability-matrix.md` does not exist.
+State-driven: reached by auto-chain from Step 4 (completed or skipped).
 
-Action: Read and execute `.cursor/skills/decompose/SKILL.md`
+Action: Read and execute `.cursor/skills/test-spec/SKILL.md`.
+
+This step converts the greenfield problem statement, acceptance criteria, solution, architecture, component docs, and UI design artifacts (if any) into test specifications before implementation begins. The test spec should cover unit, integration, blackbox, and e2e scenarios where those levels are applicable to the project.
+
+---
+
+**Step 6 — Decompose**
+Condition: `_docs/02_document/` contains `architecture.md` AND `_docs/02_document/components/` has at least one component AND `_docs/02_document/tests/traceability-matrix.md` exists AND `_docs/02_tasks/todo/` does not exist or has no implementation task files.
+
+Action: Invoke `.cursor/skills/decompose/SKILL.md` for **implementation task decomposition**. The greenfield flow selects the implementation entrypoint before handing off: Bootstrap Structure, Module Layout, Component Task Decomposition, and Cross-Task Verification.
+
+Do not invoke Blackbox Test Task Decomposition from Step 6. Test tasks are intentionally deferred to Step 9 (Decompose Tests) so the first implementation batch stays focused on product functionality and Step 8 can revise testability before test task files exist.
 
 If `_docs/02_tasks/` subfolders have some task files already, the decompose skill's resumability handles it.
 
 ---
 
-**Step 6 — Implement**
-Condition: `_docs/02_tasks/todo/` contains task files AND `_dependencies_table.md` exists AND `_docs/03_implementation/` does not contain any `implementation_report_*.md` file
+**Step 7 — Implement**
+Condition: `_docs/02_tasks/todo/` contains implementation task files AND `_dependencies_table.md` exists AND `_docs/03_implementation/` does not contain a valid product implementation report.
 
-Action: Read and execute `.cursor/skills/implement/SKILL.md`
+Action: Invoke `.cursor/skills/implement/SKILL.md` with task selection context **Product implementation**.
+
+The implement skill must run its **Product Implementation Completeness Gate** before it writes any final product implementation report. This gate compares completed product task specs, architecture/component promises, and actual source code so scaffold-only implementations cannot advance to Step 8. A final product implementation report without `_docs/03_implementation/implementation_completeness_cycle[N]_report.md` is incomplete and must not be treated as Step 7 completion.
 
 If `_docs/03_implementation/` has batch reports, the implement skill detects completed tasks and continues. The FINAL report filename is context-dependent — see implement skill documentation for naming convention.
 
+For folder fallback, **implementation task files** means task specs that are not test-only specs: exclude `*_test_infrastructure.md` and task specs whose `**Component**` or `**Epic**` identifies `Blackbox Tests`.
+
+For folder fallback, a **product implementation report** is any `_docs/03_implementation/implementation_report_*.md` file except `_docs/03_implementation/implementation_report_tests.md` and refactor reports. It is valid for greenfield progression only when:
+- the matching `_docs/03_implementation/implementation_completeness_cycle[N]_report.md` exists,
+- that completeness report does not contain unresolved `FAIL` classifications, and
+- `_docs/02_tasks/todo/` contains no pending implementation task files.
+
+If a product report exists but any of those validity checks fail, treat product implementation as incomplete and stay in Step 7.
+
 ---
 
-**Step 7 — Run Tests**
-Condition (folder fallback): `_docs/03_implementation/` contains an `implementation_report_*.md` file.
-State-driven: reached by auto-chain from Step 6.
+**Step 8 — Code Testability Revision**
+Condition (folder fallback): `_docs/03_implementation/` contains a valid product implementation report, `_docs/03_implementation/implementation_completeness_cycle[N]_report.md` exists without unresolved `FAIL` classifications, `_docs/04_refactoring/01-testability-refactoring/testability_assessment.md` does not exist, `_docs/04_refactoring/01-testability-refactoring/testability_changes_summary.md` does not exist, `_docs/03_implementation/implementation_report_tests.md` does not exist, and `_docs/02_tasks/todo/` does not contain test task files.
+State-driven: reached by auto-chain from Step 7.
+
+**Purpose**: verify the newly built code can be exercised by the planned tests before writing the test suite. Greenfield code should be testable by design; this step catches accidental hardcoded paths, singletons, direct external service construction, or other implementation choices that would make meaningful tests impossible.
+
+**Scope — MINIMAL, SURGICAL fixes**: this is not a general refactor. It is the smallest set of changes required to make the implemented code runnable under tests.
+
+**Allowed changes** in this phase:
+- Replace hardcoded URLs / file paths / credentials / magic numbers with env vars or constructor arguments.
+- Extract narrow interfaces for components that need stubbing in tests.
+- Add optional constructor parameters for dependency injection; default to the existing behavior so callers do not break.
+- Wrap global singletons in thin accessors that tests can override.
+- Split a function ONLY when necessary to stub one of its collaborators — do not split for clarity alone.
+
+**NOT allowed** in this phase (defer to a later refactor task):
+- Renaming public APIs.
+- Moving code between files unless strictly required for isolation.
+- Changing algorithms or business logic.
+- Restructuring module boundaries or rewriting layers.
+
+Action: Analyze the codebase against the test specs to determine whether the code can be tested as-is.
+
+1. Read `_docs/02_document/tests/traceability-matrix.md` and all test scenario files in `_docs/02_document/tests/`.
+2. For each test scenario, check whether the code under test can be exercised in isolation. Look for:
+   - Hardcoded file paths or directory references
+   - Hardcoded configuration values (URLs, credentials, magic numbers)
+   - Global mutable state that cannot be overridden
+   - Tight coupling to external services without abstraction
+   - Missing dependency injection or non-configurable parameters
+   - Direct file system operations without path configurability
+   - Inline construction of heavy dependencies (models, clients)
+3. If ALL scenarios are testable as-is:
+   - Create `_docs/04_refactoring/01-testability-refactoring/`
+   - Write `_docs/04_refactoring/01-testability-refactoring/testability_assessment.md` with the scenarios reviewed and outcome "Code is testable — no changes needed"
+   - Mark Step 8 as `completed` with outcome "Code is testable — no changes needed"
+   - Auto-chain to Step 9 (Decompose Tests)
+4. If testability issues are found:
+   - Create `_docs/04_refactoring/01-testability-refactoring/`
+   - Write `list-of-changes.md` in that directory using the refactor skill template (`.cursor/skills/refactor/templates/list-of-changes.md`), with:
+     - **Mode**: `guided`
+     - **Source**: `autodev-greenfield-testability-analysis`
+     - One change entry per testability issue found (change ID, file paths, problem, proposed change, risk, dependencies). Each entry must fit the allowed-changes list above; reject entries that drift into full refactor territory and log them under "Deferred refactor candidates" instead.
+   - Invoke the refactor skill in **guided mode**: read and execute `.cursor/skills/refactor/SKILL.md` with the `list-of-changes.md` as input
+   - Phase 3 (Safety Net) is skipped for this testability run because the test suite has not been implemented yet
+   - After execution, surface `RUN_DIR/testability_changes_summary.md` to the user via the Choose format (accept / request follow-up) before auto-chaining
+   - Copy or save the accepted summary as `_docs/04_refactoring/01-testability-refactoring/testability_changes_summary.md` so folder fallback can detect Step 8 completion
+   - Mark Step 8 as `completed`
+   - Auto-chain to Step 9 (Decompose Tests)
+
+---
+
+**Step 9 — Decompose Tests**
+Condition (folder fallback): `_docs/02_document/tests/traceability-matrix.md` exists AND workspace contains source code files AND `_docs/03_implementation/` contains a valid product implementation report AND `_docs/03_implementation/implementation_completeness_cycle[N]_report.md` exists without unresolved `FAIL` classifications AND (`_docs/04_refactoring/01-testability-refactoring/testability_assessment.md` exists OR `_docs/04_refactoring/01-testability-refactoring/testability_changes_summary.md` exists) AND (`_docs/02_tasks/todo/` does not exist or has no test task files) AND `_docs/03_implementation/implementation_report_tests.md` does not exist.
+State-driven: reached by auto-chain from Step 8.
+
+Action: Read and execute `.cursor/skills/decompose/SKILL.md` in **tests-only mode** (pass `_docs/02_document/tests/` as input). The decompose skill will:
+1. Run Step 1t (test infrastructure bootstrap)
+2. Run Step 3 (blackbox/e2e-capable test task decomposition)
+3. Run Step 4 (cross-verification against test coverage)
+
+If `_docs/02_tasks/` subfolders have some task files already, the decompose skill's resumability handles it — it appends test tasks alongside existing completed implementation tasks.
+
+---
+
+**Step 10 — Implement Tests**
+Condition (folder fallback): `_docs/02_tasks/todo/` contains test task files AND `_dependencies_table.md` exists AND `_docs/03_implementation/implementation_report_tests.md` does not exist.
+State-driven: reached by auto-chain from Step 9.
+
+Action: Invoke `.cursor/skills/implement/SKILL.md` with task selection context **Test implementation**.
+
+The implement skill reads only test tasks from `_docs/02_tasks/todo/` and implements them.
+
+If `_docs/03_implementation/` has batch reports, the implement skill detects completed test tasks and continues.
+
+For folder fallback, **test task files** means `*_test_infrastructure.md` plus task specs whose `**Component**` or `**Epic**` identifies `Blackbox Tests`.
+
+---
+
+**Step 11 — Run Tests**
+Condition (folder fallback): `_docs/03_implementation/implementation_report_tests.md` exists.
+State-driven: reached by auto-chain from Step 10.
 
 Action: Read and execute `.cursor/skills/test-run/SKILL.md`
 
+Verifies the implemented unit, integration, blackbox, and e2e tests pass before proceeding to spec and documentation sync. This is a hard product gate, not a harness-smoke gate: e2e/blackbox tests must exercise the actual implemented system through public runtime boundaries and compare actual outputs against `_docs/00_problem/input_data/expected_results/results_report.md` or referenced machine-readable expected-result files. Stubs are allowed only for external systems outside the product boundary; missing internal product implementation must fail or block the gate and send the flow back to Implement.
+
 ---
 
-**Step 8 — Security Audit (optional)**
-State-driven: reached by auto-chain from Step 7.
+**Step 12 — Test-Spec Sync**
+State-driven: reached by auto-chain from Step 11. Requires `_docs/02_document/tests/traceability-matrix.md` to exist — if missing, mark Step 12 `skipped` (see Action below).
+
+Action: Read and execute `.cursor/skills/test-spec/SKILL.md` in **cycle-update mode**. Pass the completed implementation task specs, completed test task specs, and implementation reports as inputs.
+
+The skill appends implementation-learned acceptance criteria, scenarios, and NFR updates to the existing test-spec files without rewriting unaffected sections. If `traceability-matrix.md` is missing, mark Step 12 as `skipped` — the next `/test-spec` full run will regenerate it.
+
+After completion, auto-chain to Step 13 (Update Docs).
+
+---
+
+**Step 13 — Update Docs**
+State-driven: reached by auto-chain from Step 12 (completed or skipped). Requires `_docs/02_document/` to contain existing documentation — if missing, mark Step 13 `skipped` (see Action below).
+
+Action: Read and execute `.cursor/skills/document/SKILL.md` in **Task mode**. Pass all completed implementation and test task spec files plus the implementation reports.
+
+The document skill in Task mode updates affected module docs, component docs, system-level docs, and test documentation without redoing full discovery, verification, or problem extraction.
+
+If `_docs/02_document/` does not contain existing docs, mark Step 13 as `skipped`.
+
+After completion, auto-chain to Step 14 (Security Audit).
+
+---
+
+**Step 14 — Security Audit (optional)**
+State-driven: reached by auto-chain from Step 13 (completed or skipped).
 
 Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Gate") with:
 - question:        `Run security audit before deploy?`
@@ -128,12 +263,12 @@ Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Ga
 - option-b-label:  `Skip — proceed directly to deploy`
 - recommendation:  `A — catches vulnerabilities before production`
 - target-skill:    `.cursor/skills/security/SKILL.md`
-- next-step:       Step 9 (Performance Test)
+- next-step:       Step 15 (Performance Test)
 
 ---
 
-**Step 9 — Performance Test (optional)**
-State-driven: reached by auto-chain from Step 8.
+**Step 15 — Performance Test (optional)**
+State-driven: reached by auto-chain from Step 14 (completed or skipped).
 
 Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Gate") with:
 - question:        `Run performance/load tests before deploy?`
@@ -141,30 +276,59 @@ Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Ga
 - option-b-label:  `Skip — proceed directly to deploy`
 - recommendation:  `A or B — base on whether acceptance criteria include latency, throughput, or load requirements`
 - target-skill:    `.cursor/skills/test-run/SKILL.md` in **perf mode** (the skill handles runner detection, threshold comparison, and its own A/B/C gate on threshold failures)
-- next-step:       Step 10 (Deploy)
+- next-step:       Step 16 (Deploy)
 
 ---
 
-**Step 10 — Deploy**
-State-driven: reached by auto-chain from Step 9 (after Step 9 is completed or skipped).
+**Step 16 — Deploy (optional)**
+State-driven: reached by auto-chain from Step 15 (after Step 15 is completed or skipped).
 
-Action: Read and execute `.cursor/skills/deploy/SKILL.md`.
+Action: Apply the **Optional Skill Gate** (`protocols.md` → "Optional Skill Gate") with:
+- question:        `Run deploy planning (scripts, procedures, compose overlays) now?`
+- option-a-label:  `Run deploy — produce/update deploy artifacts and scripts`
+- option-b-label:  `Skip — continue development; deploy when ready for production`
+- recommendation:  `B when the product is not ready to ship; A when targeting a release soon`
+- target-skill:    `.cursor/skills/deploy/SKILL.md`
+- next-step:       Step 16.5 (Release) — only when Step 16 was completed; otherwise Step 17 (Retrospective)
 
-After the deploy skill completes successfully, mark Step 10 as `completed` and auto-chain to Step 11 (Retrospective).
+On **skip**: mark Step 16 and Step 16.5 as `skipped`; record in the release report (if one exists) or `_docs/_autodev_state.md` `sub_step.detail` that deploy/release were deferred; auto-chain to Step 17 (Retrospective in cycle-end mode).
+
+On **complete**: mark Step 16 `completed` and auto-chain to Step 16.5 (Release).
 
 ---
 
-**Step 11 — Retrospective**
-State-driven: reached by auto-chain from Step 10.
+**Step 16.5 — Release (optional)**
+State-driven: reached by auto-chain from Step 16 **only when Step 16 status is `completed`**. If Step 16 was `skipped`, Step 16.5 is also `skipped` and the flow does not invoke `/release`.
 
-Action: Read and execute `.cursor/skills/retrospective/SKILL.md` in **cycle-end mode**. This closes the cycle's feedback loop by folding metrics into `_docs/06_metrics/retro_<date>.md` and appending the top-3 lessons to `_docs/LESSONS.md`.
+Action: Read and execute `.cursor/skills/release/SKILL.md`. The release skill is responsible for selecting the target environment, executing the deploy artifacts, smoke-testing, watching the rollout, and producing a definitive verdict (`Released`, `Released-with-override`, `Rolled-Back`, or `Aborted`).
 
-After retrospective completes, mark Step 11 as `completed` and enter "Done" evaluation.
+The release skill has its own internal BLOCKING gates (Phase 1 pre-release gate, Phase 2 strategy select, Phase 6 user confirmation when soft regression escalates). Autodev does NOT add a wrapping A/B/C gate — the release skill owns its own user interaction.
+
+After the release skill exits:
+
+- **Verdict `Released`** → mark Step 16.5 `completed` and auto-chain to Step 17 (Retrospective in cycle-end mode).
+- **Verdict `Released-with-override`** → mark Step 16.5 `completed` AND auto-chain to Step 17 (Retrospective in **incident mode**) — the override is itself an incident the retrospective must analyze.
+- **Verdict `Rolled-Back`** → mark Step 16.5 `failed`. Auto-chain to Step 17 (Retrospective in **incident mode**). Do NOT consider the project "Done" — the user owns the next move (re-run /implement on a fix branch, re-run /deploy, re-run /release).
+- **Verdict `Aborted`** → mark Step 16.5 `not_started` (the release was never started) OR `failed` if the abort came after Phase 3 had already touched the live system. Surface the abort reason and STOP — do not auto-chain to retrospective.
+
+---
+
+**Step 17 — Retrospective**
+State-driven: reached by auto-chain from Step 16.5 (any verdict) OR from Step 16/16.5 both `skipped` (cycle-end mode — note deploy/release deferred in the retro report).
+
+Action: Read and execute `.cursor/skills/retrospective/SKILL.md`. Mode selection:
+
+- Step 16.5 verdict `Released` → cycle-end mode
+- Step 16.5 verdict `Released-with-override` or `Rolled-Back` → incident mode
+
+The retrospective closes the cycle's feedback loop by folding metrics into `_docs/06_metrics/retro_<date>.md` (or `incident_<date>_release.md` in incident mode) and appending the top-3 lessons to `_docs/LESSONS.md`.
+
+After retrospective completes, mark Step 17 as `completed` and enter "Done" evaluation.
 
 ---
 
 **Done**
-State-driven: reached by auto-chain from Step 11. (Sanity check: `_docs/04_deploy/` should contain all expected artifacts — containerization.md, ci_cd_pipeline.md, environment_strategy.md, observability.md, deployment_procedures.md, deploy_scripts.md.)
+State-driven: reached by auto-chain from Step 17. (Sanity check: if Step 16 was `completed`, `_docs/04_deploy/` should contain the expected deploy artifacts. If Step 16.5 was `completed`, `_docs/04_release/` should contain a release report with a definitive verdict. Skipped deploy/release is valid — no release report required.)
 
 Action: Report project completion with summary. Then **rewrite the state file** so the next `/autodev` invocation enters the feature-cycle loop in the existing-code flow:
 
@@ -191,47 +355,72 @@ On the next invocation, Flow Resolution rule 1 reads `flow: existing-code` and r
 | Research (2) | Auto-chain → Research Decision (ask user: another round or proceed?) |
 | Research Decision → proceed | Auto-chain → Plan (3) |
 | Plan (3) | Auto-chain → UI Design detection (4) |
-| UI Design (4, done or skipped) | Auto-chain → Decompose (5) |
-| Decompose (5) | **Session boundary** — suggest new conversation before Implement |
-| Implement (6) | Auto-chain → Run Tests (7) |
-| Run Tests (7, all pass) | Auto-chain → Security Audit choice (8) |
-| Security Audit (8, done or skipped) | Auto-chain → Performance Test choice (9) |
-| Performance Test (9, done or skipped) | Auto-chain → Deploy (10) |
-| Deploy (10) | Auto-chain → Retrospective (11) |
-| Retrospective (11) | Report completion; rewrite state to existing-code flow, step 9 |
+| UI Design (4, done or skipped) | Auto-chain → Test Spec (5) |
+| Test Spec (5) | Auto-chain → Decompose (6) |
+| Decompose (6) | **Session boundary** — suggest new conversation before Implement |
+| Implement (7) | Auto-chain only after Product Implementation Completeness Gate passes → Code Testability Revision (8) |
+| Code Testability Revision (8) | Auto-chain → Decompose Tests (9) |
+| Decompose Tests (9) | **Session boundary** — suggest new conversation before Implement Tests |
+| Implement Tests (10) | Auto-chain → Run Tests (11) |
+| Run Tests (11, all pass) | Auto-chain → Test-Spec Sync (12) |
+| Test-Spec Sync (12, done or skipped) | Auto-chain → Update Docs (13) |
+| Update Docs (13, done or skipped) | Auto-chain → Security Audit choice (14) |
+| Security Audit (14, done or skipped) | Auto-chain → Performance Test choice (15) |
+| Performance Test (15, done or skipped) | Auto-chain → Deploy choice (16) |
+| Deploy (16, completed) | Auto-chain → Release (16.5) |
+| Deploy (16, skipped) | Mark 16.5 `skipped` → auto-chain → Retrospective (17, cycle-end mode) |
+| Release (16.5, verdict Released) | Auto-chain → Retrospective (17, cycle-end mode) |
+| Release (16.5, verdict Released-with-override) | Auto-chain → Retrospective (17, **incident mode**) |
+| Release (16.5, verdict Rolled-Back) | Auto-chain → Retrospective (17, **incident mode**); do NOT enter Done |
+| Release (16.5, verdict Aborted) | STOP — surface abort reason; do not auto-chain |
+| Retrospective (17) | Report completion; rewrite state to existing-code flow, step 9 |
 
 ## Status Summary — Step List
 
 Flow name: `greenfield`. Render using the banner template in `protocols.md` → "Banner Template (authoritative)". No header-suffix, current-suffix, or footer-extras — all empty for this flow.
 
-| # | Step Name          | Extra state tokens (beyond the shared set) |
-|---|--------------------|--------------------------------------------|
-| 1 | Problem            | — |
-| 2 | Research           | `DONE (N drafts)` |
-| 3 | Plan               | — |
-| 4 | UI Design          | — |
-| 5 | Decompose          | `DONE (N tasks)` |
-| 6 | Implement          | `IN PROGRESS (batch M of ~N)` |
-| 7 | Run Tests          | `DONE (N passed, M failed)` |
-| 8 | Security Audit     | — |
-| 9 | Performance Test   | — |
-| 10 | Deploy            | — |
-| 11 | Retrospective     | — |
+| # | Step Name                   | Extra state tokens (beyond the shared set) |
+|---|-----------------------------|--------------------------------------------|
+| 1 | Problem                     | — |
+| 2 | Research                    | `DONE (N drafts)` |
+| 3 | Plan                        | — |
+| 4 | UI Design                   | — |
+| 5 | Test Spec                   | — |
+| 6 | Decompose                   | `DONE (N tasks)` |
+| 7 | Implement                   | `IN PROGRESS (batch M of ~N)` |
+| 8 | Code Testability Revision   | — |
+| 9 | Decompose Tests             | `DONE (N tasks)` |
+| 10 | Implement Tests            | `IN PROGRESS (batch M)` |
+| 11 | Run Tests                  | `DONE (N passed, M failed)` |
+| 12 | Test-Spec Sync             | — |
+| 13 | Update Docs                | — |
+| 14 | Security Audit             | — |
+| 15 | Performance Test           | — |
+| 16 | Deploy                     | — |
+| 16.5 | Release                  | `DONE (Released | Released-with-override | Rolled-Back | Aborted)` |
+| 17 | Retrospective              | — |
 
-All rows also accept the shared state tokens (`DONE`, `IN PROGRESS`, `NOT STARTED`, `FAILED (retry N/3)`); rows 4, 8, 9 additionally accept `SKIPPED`.
+All rows also accept the shared state tokens (`DONE`, `IN PROGRESS`, `NOT STARTED`, `FAILED (retry N/3)`); rows 4, 12, 13, 14, 15, 16, 16.5 additionally accept `SKIPPED`.
 
 Row rendering format (step-number column is right-padded to 2 characters for alignment):
 
 ```
- Step 1   Problem             [<state token>]
- Step 2   Research            [<state token>]
- Step 3   Plan                [<state token>]
- Step 4   UI Design           [<state token>]
- Step 5   Decompose           [<state token>]
- Step 6   Implement           [<state token>]
- Step 7   Run Tests           [<state token>]
- Step 8   Security Audit      [<state token>]
- Step 9   Performance Test    [<state token>]
- Step 10  Deploy              [<state token>]
- Step 11  Retrospective       [<state token>]
+ Step 1   Problem                   [<state token>]
+ Step 2   Research                  [<state token>]
+ Step 3   Plan                      [<state token>]
+ Step 4   UI Design                 [<state token>]
+ Step 5   Test Spec                 [<state token>]
+ Step 6   Decompose                 [<state token>]
+ Step 7   Implement                 [<state token>]
+ Step 8   Code Testability Rev.     [<state token>]
+ Step 9   Decompose Tests           [<state token>]
+ Step 10  Implement Tests           [<state token>]
+ Step 11  Run Tests                 [<state token>]
+ Step 12  Test-Spec Sync            [<state token>]
+ Step 13  Update Docs               [<state token>]
+ Step 14  Security Audit            [<state token>]
+ Step 15  Performance Test          [<state token>]
+ Step 16  Deploy                    [<state token>]
+ Step 16.5 Release                  [<state token>]
+ Step 17  Retrospective             [<state token>]
 ```

.cursor/skills/autodev/flows/meta-repo.md

@@ -5,7 +5,8 @@ Workflow for **meta-repositories** — repos that aggregate multiple components
 This flow differs fundamentally from `greenfield` and `existing-code`:
 
 - **No problem/research/plan phases** — meta-repos don't build features, they coordinate existing ones
-- **No test spec / implement / run tests** — the meta-repo has no code to test
+- **No test spec / run tests** — the meta-repo has no code to test
+- **`implement` is scoped to suite-level work only** — cross-repo concerns, repo/folder renames, suite-root infra additions (e.g., `.gitmodules`, `_infra/`, suite `e2e/`). Per-component implementation lives in each component's own workspace `/autodev` cycle. The meta-repo's implement step (Step 3.5) executes only when `_docs/tasks/todo/` is non-empty AND the user explicitly opts in; placement is **before** the sync skills so subsequent Doc/E2E/CICD sync propagates the post-implementation state.
 - **No `_docs/00_problem/` artifacts** — documentation target is `_docs/*.md` unified docs, not per-feature `_docs/NN_feature/` folders
 - **Primary artifact is `_docs/_repo-config.yaml`** — generated by `monorepo-discover`, read by every other step
 
@@ -17,6 +18,7 @@ This flow differs fundamentally from `greenfield` and `existing-code`:
 | 2 | Config Review | (human checkpoint, no sub-skill) | — |
 | 2.5 | Glossary & Architecture Vision | (inline, no sub-skill) | Steps 1–5 |
 | 3 | Status | monorepo-status/SKILL.md | Sections 1–5 |
+| 3.5 | Suite Implement | implement/SKILL.md (suite-level invocation context) | Steps 1–14 + 16 (Step 14.5 + Step 15 skipped); conditional on `_docs/tasks/todo/` non-empty AND user opt-in |
 | 4 | Document Sync | monorepo-document/SKILL.md | Phase 1–7 (conditional on doc drift) |
 | 4.5 | Integration Test Sync | monorepo-e2e/SKILL.md | Phase 1–6 (conditional on suite-e2e drift; skipped if `suite_e2e:` block absent in config) |
 | 5 | CICD Sync | monorepo-cicd/SKILL.md | Phase 1–7 (conditional on CI drift) |
@@ -184,11 +186,16 @@ The status report identifies:
 - Registry/config mismatches
 - Unresolved questions
 
-Based on the report, auto-chain branches:
+Based on the report, auto-chain branches in this evaluation order (first match wins):
 
-- If **doc drift** found → auto-chain to **Step 4 (Document Sync)**
-- Else if **CI drift** (only) found → auto-chain to **Step 5 (CICD Sync)**
-- Else if **registry mismatch** found (new components not in config) → present Choose format:
+1. **Registry mismatch** (new components not in config, or config component not in registry) → present the Choose format below FIRST. After the user resolves it (A: refresh discover, B: onboard, C: continue with mismatch acknowledged), proceed to the next rule. This rule has priority because a stale config would mislead Step 3.5's ownership-envelope synthesis and any sync skill's component scope.
+2. **Pre-routing gate (Step 3.5 detection)** — check `_docs/tasks/todo/` for suite-level task files (`*.md` excluding files starting with `_`). If ≥1 task is present, auto-chain to **Step 3.5 (Suite Implement)**. After Step 3.5 returns (regardless of A/B outcome), the post-implement re-status applies rules 3–6 below to the post-implementation state.
+3. If **doc drift** found → auto-chain to **Step 4 (Document Sync)**
+4. Else if **CI drift** (only) found → auto-chain to **Step 5 (CICD Sync)**
+5. Else if **suite-e2e drift** (only) found → auto-chain to **Step 4.5 (Integration Test Sync)** (only when `suite_e2e:` block exists in config)
+6. Else → **workflow done for this cycle**.
+
+**Registry mismatch Choose format** (rule 1):
 
 ```
 ══════════════════════════════════════
@@ -205,7 +212,134 @@ Based on the report, auto-chain branches:
 ══════════════════════════════════════
 ```
 
-- Else → **workflow done for this cycle**. Report "No drift. Meta-repo is in sync." Loop waits for next invocation.
+When rule 6 fires (no drift, no todo tasks), report "No drift. Meta-repo is in sync." and end the cycle. Loop waits for next invocation.
+
+---
+
+**Step 3.5 — Suite Implement**
+
+Condition (folder fallback): `_docs/tasks/todo/` exists AND contains ≥1 file matching `*.md` excluding files starting with `_` (e.g., `_dependencies_table.md` is excluded by convention).
+
+State-driven: reached by auto-chain from Step 3 when the pre-routing gate detected todo tasks. Inserted **before** the sync skills (Step 4 / 4.5 / 5) by deliberate design: implementing renames + cross-repo edits first means the subsequent sync skills propagate the actual landed state rather than the pre-change state, avoiding a second cycle to fix downstream drift.
+
+**Skip condition**: `_docs/tasks/todo/` is empty, missing, or contains only `_*` files. In that case Step 3.5 is skipped entirely and the cycle proceeds with Step 3's existing drift-based routing.
+
+**Goal**: Execute suite-level implementation tasks — cross-repo concerns (e.g., `autopilot` + `ui` + suite `e2e/` cutover in a coordinated change-set), folder renames (e.g., `git mv flights missions` + `.gitmodules` edit + `_infra/` path refs), and suite-root infrastructure additions (e.g., `_infra/dev/docker-compose.dev.yml`). Per-component implementation work stays in each component's own workspace `/autodev` cycle.
+
+**Why this exists**: the meta-repo's existing sync skills (`monorepo-document`, `monorepo-cicd`, `monorepo-e2e`) only **propagate** changes that already landed. They cannot **execute** a task spec. Without Step 3.5, suite-level tickets like AZ-543 (B4 repo rename) or AZ-506 (new dev compose) have no flow path forward — they require operator action outside autodev.
+
+**Inputs**:
+
+- `_docs/tasks/todo/*.md` (excluding `_*`) — task specs in the existing format (`Task` / `Component` / `Dependencies` / `Acceptance criteria` headers)
+- `_docs/_repo-config.yaml` — `components[].path` list, used to compute the suite-level OWNED envelope (workspace root EXCLUDING any path under a component's folder)
+- `_docs/tasks/_dependencies_table.md` — synthesized by this step if missing (see Procedure)
+- `_docs/tasks/_suite_module_layout.md` — synthesized by this step if missing (see Procedure)
+
+**Procedure**:
+
+1. **Detection (already done by Step 3 pre-routing gate)**. List task files in `_docs/tasks/todo/` (excluding `_*`). If 0 → skip Step 3.5. If ≥1 → continue.
+
+2. **Present Choose**:
+
+   ```
+   ══════════════════════════════════════
+    DECISION REQUIRED: <N> suite-level task(s) in _docs/tasks/todo/
+   ══════════════════════════════════════
+    Task(s) detected:
+      - AZ-XXX: <title>           (deps: <list or "—">)
+      - AZ-YYY: <title>           (deps: <list or "—">)
+      ...
+
+    A) Run implement skill on these task(s) now (then continue to Doc / E2E / CICD sync)
+    B) Skip implement this cycle — continue to Doc / E2E / CICD sync without executing tasks
+    C) Pause — review the tasks before deciding (end session, no state changes)
+   ══════════════════════════════════════
+    Recommendation: A — running implement BEFORE syncs means subsequent
+                    sync skills propagate the post-implementation state.
+                    B is appropriate when tasks are blocked on user input
+                    or external coordination. C when the tasks themselves
+                    need owner clarification before execution.
+   ══════════════════════════════════════
+   ```
+
+3. **On user A — Pre-flight**:
+
+   a. **Working tree clean check**. Run `git status --porcelain`. If non-empty, surface to the user with a Choose A/B/C identical to the implement skill's prerequisite gate (commit/stash manually; agent commits as `chore: WIP pre-implement`; abort).
+
+   b. **Synthesize `_docs/tasks/_dependencies_table.md`** if missing. Parse each in-scope task's `Dependencies:` field. Write a minimal table of the form:
+
+      ```markdown
+      # Suite-Level Task Dependencies
+
+      | Task ID | Depends on | Notes |
+      |---------|------------|-------|
+      | AZ-XXX  | (none)     | — |
+      | AZ-YYY  | AZ-XXX     | — |
+      ```
+
+      If a task lists a dependency that is neither in `todo/` nor `done/`, log a warning in the synthesized file but do not block — implement skill's Step 1 (Parse) will surface the issue if it actually blocks execution.
+
+   c. **Synthesize `_docs/tasks/_suite_module_layout.md`** if missing. Default content:
+
+      ```markdown
+      # Suite-Level Module Layout (synthetic)
+
+      Generated by autodev meta-repo Step 3.5. The suite root has no per-feature decomposition; ownership is defined at the component-boundary level only.
+
+      ## Per-Component Mapping
+
+      | Component | Owns                             | Imports from |
+      |-----------|----------------------------------|--------------|
+      | suite     | (workspace root) excluding any path listed under `_repo-config.yaml.components[].path` | (read-only) every component's primary doc + `_docs/*.md` |
+
+      Suite-level tasks operate on: `.gitmodules`, `_infra/**`, `_docs/**` (excluding `_docs/tasks/_*` regenerated files), root `README.md`, `e2e/**` (suite e2e harness only).
+
+      Forbidden paths for suite-level tasks: `<component>/**` for every component listed in `_repo-config.yaml.components[].path` — those edits live in the component's own workspace `/autodev` cycle.
+      ```
+
+   d. **Prepare invocation context**:
+
+      ```
+      suite_level: true
+      TASKS_DIR: _docs/tasks/
+      module_layout_path: _docs/tasks/_suite_module_layout.md
+      ```
+
+4. **Invoke implement skill**. Read and execute `.cursor/skills/implement/SKILL.md` with the prepared context. The skill's "Suite-level invocation context" subsection (added in tandem with this flow change) honors the three flags above and skips:
+
+   - Step 14.5 (cumulative code review) — no `architecture_compliance_baseline.md` exists at the suite level; cross-task drift is captured by the next `monorepo-status` cycle instead.
+   - Step 15 (Product Implementation Completeness Gate) — the gate's inputs (`_docs/02_document/architecture.md`, `system-flows.md`, `components/*/description.md`) do not exist in the meta-repo artifact layout. Suite tasks are infrastructure / coordination work, not feature implementation.
+
+   All other implement skill steps (1–14, 16) execute unchanged. Tracker integration (Step 5: In Progress, Step 12: In Testing) runs normally.
+
+5. **Post-implement re-status**. After the implement skill completes (last batch committed, all originally-todo tasks moved to `_docs/tasks/done/`), silently re-run Step 3's drift detection logic — do NOT re-render the full Status report; just re-evaluate the drift signals against the post-implementation tree. Then auto-chain per the post-implementation drift findings:
+
+   - Doc drift → Step 4 (Document Sync)
+   - Suite-e2e drift only → Step 4.5
+   - CI drift only → Step 5
+   - No drift → cycle complete
+
+   Note: the post-implement re-status is exactly why Step 3.5 is placed before sync. A repo rename will typically introduce doc + CI drift; the next invocation of Step 4 / Step 5 catches it on the same cycle.
+
+6. **On user B (skip)** → mark Step 3.5 `skipped` in state file. Apply Step 3's original drift-based routing (compute from the pre-Step-3.5 Status report).
+
+7. **On user C (pause)** → end session. Update state to `step: 3.5, status: in_progress, sub_step: {phase: 0, name: awaiting-task-review, detail: "<N> tasks pending review"}`. Tell the user to invoke `/autodev` again after deciding. **Do NOT modify any files** — pre-flight has not run yet.
+
+**Self-verification** (executed before invoking implement):
+
+- [ ] Working tree is clean (or user explicitly chose B in the WIP-stash sub-Choose)
+- [ ] `_docs/tasks/_dependencies_table.md` exists (synthesized if it didn't)
+- [ ] `_docs/tasks/_suite_module_layout.md` exists (synthesized if it didn't)
+- [ ] All in-scope task files have a `Component:` field (skip + report any that don't — don't guess ownership)
+- [ ] Tracker availability gate satisfied per `protocols.md` (or `tracker: local` previously chosen)
+
+**Failure handling**:
+
+- If implement returns FAILED → standard Failure Handling (`protocols.md`): retry up to 3 times, then escalate.
+- If implement is interrupted mid-batch → next invocation re-detects via the implement skill's resumability protocol (read latest `_docs/03_implementation/suite_batch_*.md`). Step 3.5 itself is reentrant: on re-entry, if `todo/` still has tasks, it presents the Choose again with the remaining set.
+- **Half-applied state risk** (acknowledged): if implement is interrupted between commits, the working tree is clean at the last commit boundary but the in-flight batch is lost. The user is responsible for inspecting and re-invoking. This is intentional — automated rollback of suite-level renames + `.gitmodules` edits is more dangerous than a human-driven recovery.
+
+**Idempotency**: if `_docs/tasks/todo/` becomes empty after this step (all tasks moved to `done/`), the next `/autodev` invocation skips Step 3.5 entirely and proceeds with normal Status → sync flow.
 
 ---
 
@@ -287,11 +421,16 @@ After onboarding completes, the config is updated. Auto-chain back to **Step 3 (
 | Config Review (2, user picked A, confirmed_by_user: true) | Auto-chain → Glossary & Architecture Vision (2.5) |
 | Config Review (2, user picked B) | **Session boundary** — end session, await re-invocation |
 | Glossary & Architecture Vision (2.5) | Auto-chain → Status (3) |
-| Status (3, doc drift) | Auto-chain → Document Sync (4) |
-| Status (3, suite-e2e drift only) | Auto-chain → Integration Test Sync (4.5) |
-| Status (3, CI drift only) | Auto-chain → CICD Sync (5) |
-| Status (3, no drift) | **Cycle complete** — end session, await re-invocation |
+| Status (3, todo tasks present) | Auto-chain → Suite Implement (3.5) — pre-routing gate fires before drift-based routing |
+| Status (3, no todo tasks, doc drift) | Auto-chain → Document Sync (4) |
+| Status (3, no todo tasks, suite-e2e drift only) | Auto-chain → Integration Test Sync (4.5) |
+| Status (3, no todo tasks, CI drift only) | Auto-chain → CICD Sync (5) |
+| Status (3, no todo tasks, no drift) | **Cycle complete** — end session, await re-invocation |
 | Status (3, registry mismatch) | Ask user (A: discover, B: onboard, C: continue) |
+| Suite Implement (3.5, user picked A, success) | Silent re-status; auto-chain per post-implementation drift (Step 4 / 4.5 / 5 / cycle complete) |
+| Suite Implement (3.5, user picked B) | Mark `skipped`; auto-chain per Step 3's original drift findings |
+| Suite Implement (3.5, user picked C) | **Session boundary** — end session, await re-invocation |
+| Suite Implement (3.5, FAILED ×3) | Standard Failure Handling escalation (`protocols.md`) |
 | Document Sync (4) + suite-e2e drift pending | Auto-chain → Integration Test Sync (4.5) |
 | Document Sync (4) + CI drift only pending | Auto-chain → CICD Sync (5) |
 | Document Sync (4) + no further drift | **Cycle complete** |
@@ -317,11 +456,12 @@ Flow-specific slot values:
 | 2 | Config Review                      | `IN PROGRESS (awaiting human)` |
 | 2.5 | Glossary & Architecture Vision   | `SKIPPED (already captured)` |
 | 3 | Status                             | `DONE (no drift)`, `DONE (N drifts)` |
+| 3.5 | Suite Implement                  | `DONE (N tasks)`, `SKIPPED (no todo tasks)`, `SKIPPED (user picked B)`, `IN PROGRESS (batch M of ~N)`, `IN PROGRESS (awaiting-task-review)` |
 | 4 | Document Sync                      | `DONE (N docs)`, `SKIPPED (no doc drift)` |
 | 4.5 | Integration Test Sync            | `DONE (N files)`, `SKIPPED (no suite-e2e drift)`, `SKIPPED (no suite_e2e config block)` |
 | 5 | CICD Sync                          | `DONE (N files)`, `SKIPPED (no CI drift)` |
 
-All rows accept the shared state tokens (`DONE`, `IN PROGRESS`, `NOT STARTED`, `FAILED (retry N/3)`); rows 2.5, 4, 4.5, and 5 additionally accept `SKIPPED`.
+All rows accept the shared state tokens (`DONE`, `IN PROGRESS`, `NOT STARTED`, `FAILED (retry N/3)`); rows 2.5, 3.5, 4, 4.5, and 5 additionally accept `SKIPPED`.
 
 Row rendering format:
 
@@ -330,6 +470,7 @@ Row rendering format:
  Step 2     Config Review                     [<state token>]
  Step 2.5   Glossary & Architecture Vision    [<state token>]
  Step 3     Status                            [<state token>]
+ Step 3.5   Suite Implement                   [<state token>]
  Step 4     Document Sync                     [<state token>]
  Step 4.5   Integration Test Sync             [<state token>]
  Step 5     CICD Sync                         [<state token>]
@@ -337,8 +478,12 @@ Row rendering format:
 
 ## Notes for the meta-repo flow
 
-- **No session boundary except Step 2 and Step 2.5**: unlike existing-code flow (which has boundaries around decompose), meta-repo flow only pauses at config review and the one-shot glossary/vision capture. Once both are confirmed, syncing is fast enough to complete in one session and Step 2.5 idempotently no-ops on every subsequent invocation.
+- **Session boundaries**: Step 2 (Config Review pending), Step 2.5 (one-shot glossary/vision review), and Step 3.5 (when user picks C "Pause"). Step 3.5's A/B picks do NOT cross a session boundary — they auto-chain to syncs in the same session.
 - **Cyclical, not terminal**: no "done forever" state. Each invocation completes a drift cycle; next invocation starts fresh.
-- **No tracker integration**: this flow does NOT create Jira/ADO tickets. Maintenance is not a feature — if a feature-level ticket spans the meta-repo's concerns, it lives in the per-component workspace.
+- **Tracker integration scope**: this flow does NOT create Jira/ADO tickets in its sync skills (Status / Document Sync / E2E / CICD). Step 3.5 (Suite Implement) IS tracker-integrated — it transitions existing tickets In Progress → In Testing per the implement skill's standard tracker handling. Suite-level tickets are authored manually by the operator (typically as children of an Epic that spans multiple components, like AZ-539); the flow doesn't auto-create them.
+- **Per-component vs. suite-level work**:
+  - Tickets that touch component source code (`<component>/src/**`) belong in that component's own workspace `/autodev` cycle. The meta-repo flow does NOT execute them.
+  - Tickets that touch suite-root paths only (`.gitmodules`, `_infra/**`, suite `e2e/**`, root `README.md`, suite `_docs/**` outside `tasks/_*`) are eligible for Step 3.5.
+  - Tickets that span both (e.g., AZ-550 B11 consumer cutover, which touches `autopilot/`, `ui/`, AND suite `e2e/`) are NOT executable from a single workspace by design — split the ticket so the suite-level slice can run in Step 3.5 and the component slices run in their owning workspaces.
 - **Onboarding is opt-in**: never auto-onboarded. User must explicitly request.
 - **Failure handling**: uses the same retry/escalation protocol as other flows (see `protocols.md`).

.cursor/skills/autodev/protocols.md

@@ -110,9 +110,11 @@ Before entering a step from this table for the first time in a session, verify t
 | Flow | Step | Sub-Step | Tracker Action |
 |------|------|----------|----------------|
 | greenfield | Plan | Step 6 — Epics | Create epics for each component |
-| greenfield | Decompose | Step 1 + Step 2 + Step 3 — All tasks | Create ticket per task, link to epic |
+| greenfield | Decompose | Implementation decomposition Step 1 + Step 2 — Product tasks | Create ticket per product task, link to epic |
+| greenfield | Decompose Tests | Step 1t + Step 3 — All test tasks | Create ticket per task, link to epic |
 | existing-code | Decompose Tests | Step 1t + Step 3 — All test tasks | Create ticket per task, link to epic |
 | existing-code | New Task | Step 7 — Ticket | Create ticket per task, link to epic |
+| meta-repo | Suite Implement | Step 3.5 — implement skill Step 5 / Step 12 | Transition existing tickets In Progress → In Testing per implement skill (does NOT create new tickets — operator authors them) |
 
 ### State File Marker
 
@@ -387,7 +389,7 @@ The banner shell is defined here once. Each flow file contributes only its step-
   where `<state token>` comes from the state-token set defined per row in the flow's step-list table.
 - `<current-suffix>` — optional, flow-specific. The existing-code flow appends ` (cycle <N>)` when `state.cycle > 1`; other flows leave it empty.
 - `Retry:` row — omit entirely when `retry_count` is 0. Include it with `<N>/3` otherwise.
-- `<footer-extras>` — optional, flow-specific. The meta-repo flow adds a `Config:` line with `_docs/_repo-config.yaml` state; other flows leave it empty.
+- `<footer-extras>` — optional, flow-specific. The meta-repo flow adds a `Config:` line with `_docs/_repo-config.yaml` state; other flows leave it empty unless **parent suite docs** apply: if `<workspace-root>/../docs` exists and is a directory, append `Suite docs (parent): <absolute path>` on its own line (or `Suite docs (parent): absent` is **not** required — omit when missing). This line is orthogonal to flow-specific footer lines; both may appear.
 
 ### State token set (shared)
 

.cursor/skills/autodev/state.md

@@ -13,7 +13,7 @@ The autodev persists its position to `_docs/_autodev_state.md`. This is a lightw
 
 ## Current Step
 flow: [greenfield | existing-code | meta-repo]
-step: [1-11 for greenfield, 1-17 for existing-code, 1-6 for meta-repo, or "done"]
+step: [1-17 for greenfield (incl. fractional 16.5), 1-17 for existing-code (incl. fractional 16.5), 1-6 for meta-repo (incl. fractional 2.5 and 3.5), or "done"]
 name: [step name from the active flow's Step Reference Table]
 status: [not_started / in_progress / completed / skipped / failed]
 sub_step:
@@ -82,6 +82,19 @@ retry_count: 0
 cycle: 1
 ```
 
+```
+flow: meta-repo
+step: 3.5
+name: Suite Implement
+status: in_progress
+sub_step:
+  phase: 7
+  name: batch-loop
+  detail: "AZ-543 batch 1 of 1; suite-level"
+retry_count: 0
+cycle: 1
+```
+
 ```
 flow: existing-code
 step: 10
@@ -100,7 +113,7 @@ cycle: 3
 1. **Create** on the first autodev invocation (after state detection determines Step 1)
 2. **Update** after every change — this includes: batch completion, sub-step progress, step completion, session boundary, failed retry, or any meaningful state transition. The state file must always reflect the current reality.
 3. **Read** as the first action on every invocation — before folder scanning
-4. **Cross-check**: verify against actual `_docs/` folder contents. If they disagree, trust the folder structure and update the state file
+4. **Cross-check**: verify against actual `_docs/` folder contents. If they disagree, trust the folder structure and update the state file. **Parent suite `docs/`**: on every invocation, also probe `<workspace-root>/../docs` (the parent directory’s `docs` folder — typical suite-level shared documentation next to a component repo). If it exists, mention it in the Status Summary footer per `protocols.md`; use it only as supplemental reading context unless a flow step explicitly ties detection to it. It never replaces workspace `_docs/` for step detection by default.
 5. **Never delete** the state file
 6. **Retry tracking**: increment `retry_count` on each failed auto-retry; reset to `0` on success. If `retry_count` reaches 3, set `status: failed`
 7. **Failed state on re-entry**: if `status: failed` with `retry_count: 3`, do NOT auto-retry — present the issue to the user first

.cursor/skills/code-review/SKILL.md

@@ -2,7 +2,7 @@
 name: code-review
 description: |
   Multi-phase code review against task specs with structured findings output.
-  6-phase workflow: context loading, spec compliance, code quality, security quick-scan, performance scan, cross-task consistency.
+  7-phase workflow: context loading, spec compliance, code quality, security quick-scan, performance scan, cross-task consistency, architecture compliance.
   Produces a structured report with severity-ranked findings and a PASS/FAIL/PASS_WITH_WARNINGS verdict.
   Invoked by /implement skill after each batch, or manually.
   Trigger phrases:
@@ -106,11 +106,12 @@ When multiple tasks were implemented in the same batch:
 
 ## Phase 7: Architecture Compliance
 
-Verify the implemented code respects the architecture documented in `_docs/02_document/architecture.md` and the component boundaries declared in `_docs/02_document/module-layout.md`.
+Verify the implemented code respects the architecture documented in `_docs/02_document/architecture.md`, the component boundaries declared in `_docs/02_document/module-layout.md`, and the **accepted Architectural Decision Records** under `_docs/02_document/adr/`.
 
 **Inputs**:
 - `_docs/02_document/architecture.md` — layering, allowed dependencies, patterns
 - `_docs/02_document/module-layout.md` — per-component directories, Public API surface, `Imports from` lists, Allowed Dependencies table
+- `_docs/02_document/adr/` — every `Status: Accepted` ADR is an enforceable structural rule. `Status: Proposed`, `Status: Deprecated`, and `Status: Superseded` ADRs are NOT enforced (Proposed = not yet ratified; Deprecated/Superseded = a later ADR overturned it). If the directory does not exist or has only the index file, ADRs are skipped — log this skip in the report so the absence is visible.
 - The cumulative list of changed files (for per-batch invocation) or the full codebase (for baseline invocation)
 
 **Checks**:
@@ -125,6 +126,11 @@ Verify the implemented code respects the architecture documented in `_docs/02_do
 
 5. **Cross-cutting concerns not locally re-implemented**: if a file under a component directory contains logic that should live in `shared/<concern>/` (e.g., custom logging setup, config loader, error envelope), flag it. Severity: Medium. Category: Architecture.
 
+6. **ADR compliance**: for each `Status: Accepted` ADR, confirm the changed code does not contradict the ADR's `Decision`. Two failure modes are flagged:
+   - **ADR-Violation**: the changed code does the opposite of an Accepted ADR's `Decision`. Example: ADR-002 says "We will use Postgres for transactional data" and the changed code introduces a SQLite dependency for a transactional path. Severity: **Critical**. Category: Architecture. The finding cites the ADR by `NNN_<slug>` and the offending file/line.
+   - **ADR-Drift**: the changed code does something the ADR did not anticipate AND that materially affects the ADR's `Consequences` (positive or negative). Example: ADR-004 says "Event-driven cross-component comms" and a changed file introduces a new synchronous HTTP call between two components. Severity: **High**. Category: Architecture. The finding either proposes "Update ADR-NNN to acknowledge the new pattern" or "Remove the drift to align with ADR-NNN" — never silently accepts.
+   The check skips ADRs that are explicitly out of scope of the changed batch (e.g., ADR-001 about deployment pipeline when the batch only touches business-logic files). Use the ADR's `Evidence` section to determine scope: if no Evidence path overlaps with any changed file, skip the ADR for this batch.
+
 **Detection approach (per language)**:
 
 - Python: parse `import` / `from ... import` statements; optionally AST with `ast` module for reliable symbol resolution.
@@ -197,7 +203,7 @@ Produce a structured report with findings deduplicated and sorted by severity:
 
 Bug, Spec-Gap, Security, Performance, Maintainability, Style, Scope, Architecture
 
-`Architecture` findings come from Phase 7. They indicate layering violations, Public API bypasses, new cyclic dependencies, duplicate symbols, or cross-cutting concerns re-implemented locally.
+`Architecture` findings come from Phase 7. They indicate layering violations, Public API bypasses, new cyclic dependencies, duplicate symbols, cross-cutting concerns re-implemented locally, **ADR-Violation** (changed code contradicts an `Accepted` ADR's Decision — Critical), or **ADR-Drift** (changed code introduces a pattern that materially affects an `Accepted` ADR's Consequences without superseding it — High).
 
 ## Verdict Logic
 
@@ -232,7 +238,7 @@ The implement skill invokes code-review by:
 
 1. Reading `.cursor/skills/code-review/SKILL.md`
 2. Providing the inputs above as context (read the files, pass content to the review phases)
-3. Executing all 6 phases sequentially
+3. Executing all 7 phases sequentially
 4. Consuming the verdict from the output
 
 ### Outputs (returned to the implement skill)

.cursor/skills/decompose/SKILL.md

@@ -2,8 +2,8 @@
 name: decompose
 description: |
   Decompose planned components into atomic implementable tasks with bootstrap structure plan.
-  4-step workflow: bootstrap structure plan, component task decomposition, blackbox test task decomposition, and cross-task verification.
-  Supports full decomposition (_docs/ structure), single component mode, and tests-only mode.
+  Workflow entrypoints: implementation task decomposition, single component decomposition, and tests-only decomposition.
+  The invoking flow decides which entrypoint to run; this skill executes that selected sequence.
   Trigger phrases:
   - "decompose", "decompose features", "feature decomposition"
   - "task decomposition", "break down components"
@@ -20,7 +20,7 @@ Decompose planned components into atomic, implementable task specs with a bootst
 
 ## Core Principles
 
-- **Atomic tasks**: each task does one thing; if it exceeds 8 complexity points, split it
+- **Atomic tasks**: each task does one thing; if it exceeds 5 complexity points, split it
 - **Behavioral specs, not implementation plans**: describe what the system should do, not how to build it
 - **Flat structure**: all tasks are tracker-ID-prefixed files in TASKS_DIR — no component subdirectories
 - **Save immediately**: write artifacts to disk after each task; never accumulate unsaved work
@@ -30,14 +30,15 @@ Decompose planned components into atomic, implementable task specs with a bootst
 
 ## Context Resolution
 
-Determine the operating mode based on invocation before any other logic runs.
+Resolve the selected entrypoint from the invocation context before any other logic runs. The caller decides whether this is implementation, single component, or tests-only decomposition; this skill only executes the selected sequence.
 
-**Default** (no explicit input file provided):
+**Implementation task decomposition** (default; selected by flows before invoking this skill):
 
 - DOCUMENT_DIR: `_docs/02_document/`
 - TASKS_DIR: `_docs/02_tasks/`
 - TASKS_TODO: `_docs/02_tasks/todo/`
 - Reads from: `_docs/00_problem/`, `_docs/01_solution/`, DOCUMENT_DIR
+- Produces only implementation tasks. Blackbox/e2e test task files are produced only when the invoking flow selects tests-only decomposition.
 
 **Single component mode** (provided file is within `_docs/02_document/` and inside a `components/` subdirectory):
 
@@ -55,24 +56,25 @@ Determine the operating mode based on invocation before any other logic runs.
 - TESTS_DIR: `DOCUMENT_DIR/tests/`
 - Reads from: `_docs/00_problem/`, `_docs/01_solution/`, TESTS_DIR
 
-Announce the detected mode and resolved paths to the user before proceeding.
+Announce the selected entrypoint and resolved paths to the user before proceeding.
 
 ### Step Applicability by Mode
 
-| Step | File | Default | Single | Tests-only |
-|------|------|:-------:|:------:|:----------:|
+| Step | File | Implementation | Single | Tests-only |
+|------|------|:--------------:|:------:|:----------:|
 | 1 Bootstrap Structure | `steps/01_bootstrap-structure.md` | ✓ | — | — |
 | 1t Test Infrastructure | `steps/01t_test-infrastructure.md` | — | — | ✓ |
 | 1.5 Module Layout | `steps/01-5_module-layout.md` | ✓ | — | — |
+| 1.7 System-Pipeline Tasks | `steps/01-7_system-pipeline-tasks.md` | ✓ | — | — |
 | 2 Task Decomposition | `steps/02_task-decomposition.md` | ✓ | ✓ | — |
-| 3 Blackbox Test Tasks | `steps/03_blackbox-test-decomposition.md` | ✓ | — | ✓ |
+| 3 Blackbox Test Tasks | `steps/03_blackbox-test-decomposition.md` | — | — | ✓ |
 | 4 Cross-Verification | `steps/04_cross-verification.md` | ✓ | — | ✓ |
 
 ## Input Specification
 
 ### Required Files
 
-**Default:**
+**Implementation task decomposition:**
 
 | File | Purpose |
 |------|---------|
@@ -84,7 +86,7 @@ Announce the detected mode and resolved paths to the user before proceeding.
 | `DOCUMENT_DIR/glossary.md` | Project terminology (confirmed by user in plan Phase 2a.0 or document Step 4.5). Use it to keep task names, component references, and AC wording consistent with the user's vocabulary |
 | `DOCUMENT_DIR/system-flows.md` | System flows from plan skill |
 | `DOCUMENT_DIR/components/[##]_[name]/description.md` | Component specs from plan skill |
-| `DOCUMENT_DIR/tests/` | Blackbox test specs from plan skill |
+| `DOCUMENT_DIR/tests/` | Optional product acceptance context from test-spec skill; do not create test task files from it in this entrypoint |
 
 **Single component mode:**
 
@@ -111,7 +113,7 @@ Announce the detected mode and resolved paths to the user before proceeding.
 
 ### Prerequisite Checks (BLOCKING)
 
-**Default:**
+**Implementation task decomposition:**
 
 1. DOCUMENT_DIR contains `architecture.md` and `components/` — **STOP if missing**
 2. Create TASKS_DIR and TASKS_TODO if they do not exist
@@ -145,6 +147,8 @@ TASKS_DIR/
 
 **Naming convention**: Each task file is initially saved in `TASKS_TODO/` with a temporary numeric prefix (`[##]_[short_name].md`). After creating the work item ticket, rename the file to use the work item ticket ID as prefix (`[TRACKER-ID]_[short_name].md`). For example: `todo/01_initial_structure.md` → `todo/AZ-42_initial_structure.md`.
 
+If tracker availability fails, follow `.cursor/rules/tracker.mdc` before continuing. Only when the user explicitly chooses `tracker: local` may the numeric prefix remain; in that mode set `Tracker: pending` and `Epic: pending` in the task header and keep the task eligible for later tracker sync.
+
 ### Save Timing
 
 | Step | Save immediately after | Filename |
@@ -166,11 +170,11 @@ If TASKS_DIR subfolders already contain task files:
 
 ## Progress Tracking
 
-At the start of execution, create a TodoWrite with all applicable steps for the detected mode (see Step Applicability table). Update status as each step/component completes.
+At the start of execution, create a TodoWrite with all applicable steps for the selected entrypoint (see Step Applicability table). Update status as each step/component completes.
 
 ## Workflow
 
-### Step 1: Bootstrap Structure Plan (default mode only)
+### Step 1: Bootstrap Structure Plan (implementation mode only)
 
 Read and follow `steps/01_bootstrap-structure.md`.
 
@@ -182,25 +186,39 @@ Read and follow `steps/01t_test-infrastructure.md`.
 
 ---
 
-### Step 1.5: Module Layout (default mode only)
+### Step 1.5: Module Layout (implementation mode only)
 
 Read and follow `steps/01-5_module-layout.md`.
 
 ---
 
-### Step 2: Task Decomposition (default and single component modes)
+### Step 1.7: System-Pipeline Tasks (implementation mode only)
+
+Read and follow `steps/01-7_system-pipeline-tasks.md`.
+
+This step exists because per-component task decomposition (Step 2)
+produces one task per component but NEVER produces a task whose
+deliverable is "the production code that drives the end-to-end
+pipeline by calling each component in order against real inputs".
+The architecture document describes the loop; nobody owns it. The
+GPS-passthrough incident (May 2026) is the canonical failure this
+step prevents.
+
+---
+
+### Step 2: Task Decomposition (implementation and single component modes)
 
 Read and follow `steps/02_task-decomposition.md`.
 
 ---
 
-### Step 3: Blackbox Test Task Decomposition (default and tests-only modes)
+### Step 3: Blackbox Test Task Decomposition (tests-only mode only)
 
 Read and follow `steps/03_blackbox-test-decomposition.md`.
 
 ---
 
-### Step 4: Cross-Task Verification (default and tests-only modes)
+### Step 4: Cross-Task Verification (implementation and tests-only modes)
 
 Read and follow `steps/04_cross-verification.md`.
 
@@ -208,7 +226,7 @@ Read and follow `steps/04_cross-verification.md`.
 
 - **Coding during decomposition**: this workflow produces specs, never code
 - **Over-splitting**: don't create many tasks if the component is simple — 1 task is fine
-- **Tasks exceeding 8 points**: split them; no task should be too complex for a single implementer
+- **Tasks exceeding 5 points**: split them; no task should be too complex for a single implementer
 - **Cross-component tasks**: each task belongs to exactly one component
 - **Skipping BLOCKING gates**: never proceed past a BLOCKING marker without user confirmation
 - **Creating git branches**: branch creation is an implementation concern, not a decomposition one
@@ -221,7 +239,7 @@ Read and follow `steps/04_cross-verification.md`.
 | Situation | Action |
 |-----------|--------|
 | Ambiguous component boundaries | ASK user |
-| Task complexity exceeds 8 points after splitting | ASK user |
+| Task complexity exceeds 5 points after splitting | ASK user |
 | Missing component specs in DOCUMENT_DIR | ASK user |
 | Cross-component dependency conflict | ASK user |
 | Tracker epic not found for a component | ASK user for Epic ID |
@@ -233,15 +251,16 @@ Read and follow `steps/04_cross-verification.md`.
 ┌────────────────────────────────────────────────────────────────┐
 │          Task Decomposition (Multi-Mode)                        │
 ├────────────────────────────────────────────────────────────────┤
-│ CONTEXT: Resolve mode (default / single component / tests-only) │
+│ CONTEXT: Invoke the selected entrypoint (implementation / single / tests-only) │
 │                                                                 │
-│ DEFAULT MODE:                                                   │
+│ IMPLEMENTATION TASK DECOMPOSITION:                              │
 │  1.   Bootstrap Structure → steps/01_bootstrap-structure.md     │
 │       [BLOCKING: user confirms structure]                       │
 │  1.5  Module Layout       → steps/01-5_module-layout.md         │
 │       [BLOCKING: user confirms layout]                          │
+│  1.7  System-Pipeline     → steps/01-7_system-pipeline-tasks.md │
+│       [BLOCKING: user confirms pipeline owners]                 │
 │  2.   Component Tasks     → steps/02_task-decomposition.md      │
-│  3.   Blackbox Tests      → steps/03_blackbox-test-decomposition.md │
 │  4.   Cross-Verification  → steps/04_cross-verification.md      │
 │       [BLOCKING: user confirms dependencies]                    │
 │                                                                 │

.cursor/skills/decompose/steps/01-5_module-layout.md

@@ -16,7 +16,8 @@
 3. Each component owns ONE top-level directory. Shared code goes under `<root>/shared/` (or language equivalent).
 4. Public API surface = files in the layout's `public:` list for each component; everything else is internal and MUST NOT be imported from other components.
 5. Cross-cutting concerns (logging, error handling, config, telemetry, auth middleware, feature flags, i18n) each get ONE entry under Shared / Cross-Cutting; per-component tasks consume them (see Step 2 cross-cutting rule).
-6. Write `_docs/02_document/module-layout.md` using `templates/module-layout.md` format.
+6. **ADR cross-check**: if `_docs/02_document/adr/` exists, read every `Status: Accepted` ADR. For each, confirm the proposed module layout does not contradict the ADR's `Decision` (e.g., an ADR mandating an event-bus boundary between two components must show up as a `Imports from` exclusion in the layout; an ADR locking a layering style must show up in the Layering table). If an ADR conflicts with the language-conventional layout from step 2, the ADR wins — record the conflict in a `## ADR-driven exceptions to the conventional layout` section of `module-layout.md` with `See ADR NNN_<slug>` references. If the ADR conflict is irreconcilable (the ADR demands something the language genuinely cannot express), STOP and ask the user A/B/C: (A) update the ADR via plan Step 4.5 supersede flow, (B) accept a layered exception with documented rationale, (C) re-open architecture.
+7. Write `_docs/02_document/module-layout.md` using `templates/module-layout.md` format. Each Per-Component Mapping entry that is governed by an ADR includes a trailing `> See ADR NNN_<slug>` line.
 
 ## Self-verification
 
@@ -26,6 +27,8 @@
 - [ ] No component's `Imports from` list points at a higher layer
 - [ ] Paths follow the detected language's convention
 - [ ] No two components own overlapping paths
+- [ ] If `_docs/02_document/adr/` exists with Accepted ADRs, every layout decision that an ADR governs has a trailing `> See ADR NNN_<slug>` reference
+- [ ] No Accepted ADR is contradicted by the layout without a documented exception
 
 ## Save action
 

.cursor/skills/decompose/steps/01-7_system-pipeline-tasks.md

@@ -0,0 +1,72 @@
+# Step 1.7: System-Pipeline Tasks (implementation mode only)
+
+**Role**: Professional software architect, integration-focused.
+**Goal**: For every end-to-end pipeline named in `_docs/02_document/architecture.md` and `_docs/02_document/system-flows.md`, ensure there is exactly ONE explicit task that owns the production code that drives that pipeline against real inputs. This step prevents the failure mode where every individual component is "complete" but no production code wires them together (May 2026 GPS-passthrough incident — see `meta-rule.mdc` "When a test reveals missing production code").
+
+**Constraints**:
+
+- This step produces *integration* tasks, not per-component tasks. Per-component tasks come from Step 2.
+- An integration task's owner is typically the composition root, runtime root, main loop, or whichever component the module layout (Step 1.5) names as the "system spine". It is NEVER a leaf component.
+- Each integration task must be sized at 5 points or fewer. If the pipeline is too large for one task, split it into per-stage integration tasks (e.g. "wire ingress → C1", then "wire C1 → C5") rather than one giant task.
+
+## Inputs
+
+| File | Purpose |
+|------|---------|
+| `_docs/02_document/architecture.md` | Source of named end-to-end pipelines and their component sequences |
+| `_docs/02_document/system-flows.md` | Source of operational flows (per-frame loop, request lifecycle, batch job, etc.) |
+| `_docs/02_document/module-layout.md` | Produced by Step 1.5. Names the "system spine" component(s) — typically `runtime_root`, `app`, `main`, `composition`, or equivalent. |
+| `_docs/02_document/components/*/description.md` | Per-component contracts so you can tell which side of a seam each method lives on |
+
+## Steps
+
+1. **Enumerate end-to-end pipelines.** Read `architecture.md` and `system-flows.md`. For each named pipeline / flow that spans 2+ components, record:
+   - The pipeline name (e.g. "per-frame nav loop", "tile-cache build", "operator pre-flight verification").
+   - The ordered sequence of components it touches (e.g. `frame_source → c1_vio → c2_vpr → ... → c5_state → replay_sink`).
+   - The trigger (per-frame, per-request, scheduled, manual).
+   - The output (what the pipeline emits and to whom).
+2. **For each pipeline, locate the owner.** Use `module-layout.md` to find the component that owns the orchestration (the "spine"). If `module-layout.md` does not name one, STOP and ASK the user which component owns the pipeline. Do NOT silently default to the bootstrap structure task — bootstrap is about project skeleton, not behavior.
+3. **Check whether the pipeline is already covered by an existing task spec or by the bootstrap-structure task.** A pipeline is "covered" only if:
+   - A task spec's `Outcome` or `Acceptance Criteria` section explicitly names "drives the {pipeline_name} end-to-end against real production components", AND
+   - That task's owned files include the orchestration code (typically the spine component's main loop / entrypoint).
+4. **For every uncovered pipeline, create a system-integration task spec** in `_docs/02_tasks/todo/` using `.cursor/skills/decompose/templates/task.md`:
+   - **Component**: the spine component from step 2 (e.g. `runtime_root`).
+   - **Outcome**: the production callsite that drives the pipeline exists and runs end-to-end on real inputs.
+   - **Scope / Included**: the orchestration code (loop body, dispatcher, scheduler, entrypoint); explicit list of every component it must call in order; the data type at each seam.
+   - **Acceptance Criteria** (write each as testable):
+     - At least one production caller of every component method in the pipeline can be found by grep — name the methods explicitly.
+     - The orchestration runs against the real production component instances (NOT mocks, NOT a passthrough that bypasses them).
+     - At least one integration test exercises the orchestration end-to-end against real inputs.
+   - **Dependencies**: every per-component task whose component appears in the pipeline.
+   - **Complexity points**: ≤5; split the pipeline if it doesn't fit.
+   - **Tracker**: create a ticket immediately (per `decompose/SKILL.md` "Tracker inline" principle); rename the file to `[TRACKER-ID]_pipeline_<name>.md`.
+5. **Mark the integration task as `Dependencies` for the integration test task.** If `tests-only` decomposition has already produced an e2e/integration test task for this pipeline, append the new integration task to its `Dependencies` field so the test cannot be "made green" before the integration ships.
+
+## Anti-patterns this step explicitly blocks
+
+- **"compose_root returns a wired runtime"** prose interpreted as "the loop exists". Composition assembles the graph; it is NOT the loop. The loop is the code that pulls inputs, drives each node, and emits outputs. If grep finds zero callers of the leaf components, the loop does not exist regardless of what compose_root does.
+- **Treating the bootstrap-structure task as the home of the main loop.** Bootstrap is project skeleton (package layout, CLI scaffold, build files). It is NOT the main loop. Main loop is its own task.
+- **Per-component tasks claiming integration scope.** A C1 VIO task's deliverable is "C1 works in isolation against unit tests". A C1 task's acceptance criteria MUST NOT include "C1 is wired into the runtime" — that's the integration task's job.
+
+## Self-verification
+
+- [ ] Every pipeline named in `architecture.md` / `system-flows.md` is listed in your enumeration.
+- [ ] Every enumerated pipeline either (a) has an existing covered task, or (b) has a new integration task in `todo/`.
+- [ ] No integration task exceeds 5 complexity points.
+- [ ] Every integration task names every component in the pipeline as a `Dependencies` entry.
+- [ ] No integration task is owned by a leaf component — every owner is named in `module-layout.md` as a spine / orchestrator.
+- [ ] Every integration task has a tracker ticket created and the filename renamed to `[TRACKER-ID]_pipeline_<name>.md`.
+
+## Save action
+
+Write the new integration task files into `_docs/02_tasks/todo/`. They will be picked up by Step 2 (Task Decomposition's dependency-table writer) and by Step 4 (Cross-Verification).
+
+## Blocking
+
+**BLOCKING**: Present the pipeline enumeration + the list of new integration tasks to the user. Do NOT proceed to Step 2 until the user confirms:
+
+- The enumeration matches what they expect from the architecture documents.
+- Every uncovered pipeline now has an integration task.
+- The chosen spine owners are correct.
+
+If the user identifies a pipeline you missed, add it before proceeding. If the user names a different spine owner, update the task and re-run self-verification.

.cursor/skills/decompose/steps/02_task-decomposition.md

@@ -26,7 +26,7 @@ For each component (or the single provided component):
 4. Do not create tasks for other components — only tasks for the current component
 5. Each task should be atomic, containing 1 API or a list of semantically connected APIs
 6. Write each task spec using `templates/task.md`
-7. Estimate complexity per task (1, 2, 3, 5, 8 points); no task should exceed 8 points — split if it does
+7. Estimate complexity per task (1, 2, 3, 5 points); no task should exceed 5 points — split if it does
 8. Note task dependencies (referencing tracker IDs of already-created dependency tasks, e.g., `AZ-42_initial_structure`)
 9. **Cross-cutting rule**: if a concern spans ≥2 components (logging, config loading, auth/authZ, error envelope, telemetry, feature flags, i18n), create ONE shared task under the cross-cutting epic. Per-component tasks declare it as a dependency and consume it; they MUST NOT re-implement it locally. Duplicate local implementations are an `Architecture` finding (High) in code-review Phase 7 and a `Maintainability` finding in Phase 6.
 10. **Shared-models / shared-API rule**: classify the task as shared if ANY of the following is true:
@@ -43,16 +43,32 @@ For each component (or the single provided component):
     Consumers read the contract file, not the producer's task spec. This prevents interface drift when the producer's implementation detail leaks into consumers.
 11. **Immediately after writing each task file**: create a work item ticket, link it to the component's epic, write the work item ticket ID and Epic ID back into the task header, then rename the file from `todo/[##]_[short_name].md` to `todo/[TRACKER-ID]_[short_name].md`.
 
+## Runtime Completeness Decomposition Gate
+
+Before Step 2 is considered complete, scan `architecture.md`, `system-flows.md`, component descriptions, and the solution for named internal runtime capabilities and dependencies. Examples include BASALT/OpenVINS/Kimera, FAISS, DINOv2, ONNX/TensorRT, ALIKED/DISK, LightGlue, RANSAC, PostGIS, MAVLink emission, FDR rollover, and any "A-Z" user-visible pipeline.
+
+For every named internal capability:
+
+1. Ensure at least one implementation task explicitly owns the production integration or production algorithm.
+2. Do not treat "define protocol", "create adapter boundary", "add deterministic fallback", "create scaffold", or "prepare native bridge" as implementation of the capability unless the architecture explicitly says the real capability is out of scope.
+3. If a capability needs external hardware/data to verify, still create the production implementation task. Verification may be hardware-gated later; implementation must not be omitted.
+4. Add a `## Runtime Completeness` section to any affected task with:
+   - named capability/dependency,
+   - production code that must exist,
+   - allowed external stubs, if any,
+   - unacceptable substitutes such as fake/deterministic/internal stubs.
+
 ## Self-verification (per component)
 
 - [ ] Every task is atomic (single concern)
-- [ ] No task exceeds 8 complexity points
+- [ ] No task exceeds 5 complexity points
 - [ ] Task dependencies reference correct tracker IDs
 - [ ] Tasks cover all interfaces defined in the component spec
 - [ ] No tasks duplicate work from other components
 - [ ] Every task has a work item ticket linked to the correct epic
 - [ ] Every shared-models / shared-API task has a contract file at `_docs/02_document/contracts/<component>/<name>.md` and a `## Contract` section linking to it
 - [ ] Every cross-cutting concern appears exactly once as a shared task, not N per-component copies
+- [ ] Every named internal runtime capability has a production implementation task, not only an interface/scaffold/fallback task
 
 ## Save action
 

.cursor/skills/decompose/steps/03_blackbox-test-decomposition.md

@@ -1,4 +1,4 @@
-# Step 3: Blackbox Test Task Decomposition (default and tests-only modes)
+# Step 3: Blackbox Test Task Decomposition (tests-only mode only)
 
 **Role**: Professional Quality Assurance Engineer
 **Goal**: Decompose blackbox test specs into atomic, implementable task specs.
@@ -6,7 +6,6 @@
 
 ## Numbering
 
-- In default mode: continue sequential numbering from where Step 2 left off.
 - In tests-only mode: start from 02 (01 is the test infrastructure bootstrap from Step 1t).
 
 ## Steps
@@ -14,21 +13,26 @@
 1. Read all test specs from `DOCUMENT_DIR/tests/` (`blackbox-tests.md`, `performance-tests.md`, `resilience-tests.md`, `security-tests.md`, `resource-limit-tests.md`)
 2. Group related test scenarios into atomic tasks (e.g., one task per test category or per component under test)
 3. Each task should reference the specific test scenarios it implements and the environment/test-data specs
-4. Dependencies:
-   - In default mode: blackbox test tasks depend on the component implementation tasks they exercise
+4. Add a **System Under Test Boundary** section to every e2e/blackbox test task:
+   - The test must drive the product through public runtime boundaries and compare actual outputs to `_docs/00_problem/input_data/expected_results/results_report.md` and any referenced machine-readable expected-result files.
+   - Stubs are allowed only for external systems outside the product boundary: flight controller/SITL, QGC observer, satellite-provider/Suite service, physical Jetson hardware, physical camera, licensed public datasets, and network services.
+   - Stubs, fakes, deterministic fallbacks, monkeypatches, or direct imports are not allowed for internal product modules that the scenario is meant to validate, such as VIO, safety/anchor wrapper, satellite retrieval, anchor verification, tile manager, MAVLink output adapter, or FDR.
+   - If an internal module is not implemented, the test must fail/block as missing product implementation; it must not pass by replacing that module with a test stub.
+5. Dependencies:
    - In tests-only mode: blackbox test tasks depend on the test infrastructure bootstrap task (Step 1t)
-5. Write each task spec using `templates/task.md`
-6. Estimate complexity per task (1, 2, 3, 5, 8 points); no task should exceed 8 points — split if it does
-7. Note task dependencies (referencing tracker IDs of already-created dependency tasks)
-8. **Immediately after writing each task file**: create a work item ticket under the "Blackbox Tests" epic, write the work item ticket ID and Epic ID back into the task header, then rename the file from `todo/[##]_[short_name].md` to `todo/[TRACKER-ID]_[short_name].md`.
+6. Write each task spec using `templates/task.md`
+7. Estimate complexity per task (1, 2, 3, 5 points); no task should exceed 5 points — split if it does
+8. Note task dependencies (referencing tracker IDs of already-created dependency tasks)
+9. **Immediately after writing each task file**: create a work item ticket under the "Blackbox Tests" epic, write the work item ticket ID and Epic ID back into the task header, then rename the file from `todo/[##]_[short_name].md` to `todo/[TRACKER-ID]_[short_name].md`.
 
 ## Self-verification
 
 - [ ] Every scenario from `tests/blackbox-tests.md` is covered by a task
 - [ ] Every scenario from `tests/performance-tests.md`, `tests/resilience-tests.md`, `tests/security-tests.md`, and `tests/resource-limit-tests.md` is covered by a task
-- [ ] No task exceeds 8 complexity points
-- [ ] Dependencies correctly reference the dependency tasks (component tasks in default mode, test infrastructure in tests-only mode)
+- [ ] No task exceeds 5 complexity points
+- [ ] Dependencies correctly reference the test infrastructure task
 - [ ] Every task has a work item ticket linked to the "Blackbox Tests" epic
+- [ ] Every e2e/blackbox task forbids internal product stubs/fakes and requires comparison against expected-results artifacts
 
 ## Save action
 

.cursor/skills/decompose/steps/04_cross-verification.md

@@ -1,4 +1,4 @@
-# Step 4: Cross-Task Verification (default and tests-only modes)
+# Step 4: Cross-Task Verification (implementation and tests-only modes)
 
 **Role**: Professional software architect and analyst
 **Goal**: Verify task consistency and produce `_dependencies_table.md`.
@@ -8,17 +8,20 @@
 
 1. Verify task dependencies across all tasks are consistent
 2. Check no gaps:
-   - In default mode: every interface in `architecture.md` has tasks covering it
+   - In implementation mode: every product interface in `architecture.md` has implementation task coverage
    - In tests-only mode: every test scenario in `traceability-matrix.md` is covered by a task
+   - In implementation mode: every named internal runtime capability/dependency from architecture, solution, system flows, and component descriptions has a production implementation task, not only an interface/scaffold/fallback task
+   - In tests-only mode: every e2e/blackbox task has a System Under Test Boundary section that forbids stubbing internal product modules and requires comparison to expected-results artifacts
 3. Check no overlaps: tasks don't duplicate work
 4. Check no circular dependencies in the task graph
 5. Produce `_dependencies_table.md` using `templates/dependencies-table.md`
 
 ## Self-verification
 
-### Default mode
+### Implementation mode
 
-- [ ] Every architecture interface is covered by at least one task
+- [ ] Every product interface in `architecture.md` is covered by at least one implementation task
+- [ ] Every named internal runtime capability has a production implementation task
 - [ ] No circular dependencies in the task graph
 - [ ] Cross-component dependencies are explicitly noted in affected task specs
 - [ ] `_dependencies_table.md` contains every task with correct dependencies
@@ -26,6 +29,7 @@
 ### Tests-only mode
 
 - [ ] Every test scenario from `traceability-matrix.md` "Covered" entries has a corresponding task
+- [ ] Every e2e/blackbox task validates actual product behavior and allows stubs only for external systems
 - [ ] No circular dependencies in the task graph
 - [ ] Test task dependencies reference the test infrastructure bootstrap
 - [ ] `_dependencies_table.md` contains every task with correct dependencies

.cursor/skills/decompose/templates/dependencies-table.md

@@ -28,4 +28,4 @@ Use this template after cross-task verification. Save as `TASKS_DIR/_dependencie
 - Dependencies column lists tracker IDs (e.g., "AZ-43, AZ-44") or "None"
 - No circular dependencies allowed
 - Tasks should be listed in recommended execution order
-- The `/implement` skill reads this table to compute parallel batches
+- The `/implement` skill reads this table to compute dependency-aware batches; task execution remains sequential

.cursor/skills/decompose/templates/task.md

@@ -11,7 +11,7 @@ Save as `TASKS_DIR/[##]_[short_name].md` initially, then rename to `TASKS_DIR/[T
 **Task**: [TRACKER-ID]_[short_name]
 **Name**: [short human name]
 **Description**: [one-line description of what this task delivers]
-**Complexity**: [1|2|3|5|8] points
+**Complexity**: [1|2|3|5] points
 **Dependencies**: [AZ-43_shared_models, AZ-44_db_migrations] or "None"
 **Component**: [component name for context]
 **Tracker**: [TASK-ID]
@@ -102,8 +102,7 @@ Consumers MUST read that file — not this task spec — to discover the interfa
 - 2 points: Non-trivial, low complexity, minimal coordination
 - 3 points: Multi-step, moderate complexity, potential alignment needed
 - 5 points: Difficult, interconnected logic, medium-high risk
-- 8 points: High difficulty, high ambiguity or coordination, multiple components
-- 13 points: Too complex — split into smaller tasks
+- 8+ points: Too complex — split into smaller tasks
 
 ## Output Guidelines
 

.cursor/skills/deploy/templates/ci_cd_pipeline.md

@@ -29,7 +29,7 @@ Save as `_docs/04_deploy/ci_cd_pipeline.md`.
 ### Test
 - Unit tests: [framework and command]
 - Blackbox tests: [framework and command, uses docker-compose.test.yml]
-- Coverage threshold: 75% overall, 90% critical paths
+- Coverage threshold: 75% overall, 90% critical-path floor (100% aim) — per `.cursor/rules/cursor-meta.mdc` Quality Thresholds
 - Coverage report published as pipeline artifact
 
 ### Security

.cursor/skills/implement/SKILL.md

@@ -25,6 +25,8 @@ For each task the main agent receives a task spec, analyzes the codebase, implem
 - **Dependency-aware ordering**: tasks run only when all their dependencies are satisfied
 - **Batching for review, not parallelism**: tasks are grouped into batches so `/code-review` and commits operate on a coherent unit of work — all tasks inside a batch are still implemented one after the other
 - **Integrated review**: `/code-review` skill runs automatically after each batch
+- **Completeness before testing**: product implementation is not done until code is checked against task outcomes, included scope, architecture/component promises, named runtime dependencies, and unresolved scaffold/native placeholders — not just task AC tests
+- **Runtime dependency reality**: production code cannot satisfy a task by exposing only a protocol, fake runner, deterministic fallback, or "native bridge" placeholder when the task/architecture promises a concrete internal capability such as BASALT VIO, FAISS retrieval, LightGlue matching, or a full A-Z localization pipeline. Stubs are allowed only for external systems and tests.
 - **Auto-start**: batches start immediately — no user confirmation before a batch
 - **Gate on failure**: user confirmation is required only when code review returns FAIL
 - **Commit per batch**: after each batch is confirmed, commit. Ask the user whether to push to remote unless the user previously opted into auto-push for this session.
@@ -32,9 +34,26 @@ For each task the main agent receives a task spec, analyzes the codebase, implem
 ## Context Resolution
 
 - TASKS_DIR: `_docs/02_tasks/`
-- Task files: all `*.md` files in `TASKS_DIR/todo/` (excluding files starting with `_`)
+- Task files: selected `*.md` files in `TASKS_DIR/todo/` (excluding files starting with `_`)
 - Dependency table: `TASKS_DIR/_dependencies_table.md`
 
+### Task Selection Context
+
+The invoking flow decides which task category this run should execute. The implement skill must honor that selected context instead of consuming every file in `todo/`.
+
+| Context | Selected task files |
+|---------|---------------------|
+| Product implementation | Task specs that are not test-only and not refactoring specs |
+| Test implementation | `*_test_infrastructure.md` plus task specs whose `Component` or `Epic` identifies `Blackbox Tests` |
+| Refactoring | Task specs whose filename or task ID includes `_refactor_` |
+
+If no explicit context is provided, infer it from the active autodev step:
+- greenfield Step 7 or existing-code Step 10 → Product implementation
+- greenfield Step 10 or existing-code Step 6 → Test implementation
+- refactor Phase 4 → Refactoring
+
+Unselected task files remain in `TASKS_DIR/todo/` for their later flow step.
+
 ### Task Lifecycle Folders
 
 ```
@@ -45,9 +64,31 @@ TASKS_DIR/
 └── done/        ← completed tasks (moved here after implementation)
 ```
 
+### Suite-level invocation context (meta-repo flow)
+
+When invoked from `.cursor/skills/autodev/flows/meta-repo.md` Step 3.5 (or any caller that supplies the same context envelope), the skill receives:
+
+```
+suite_level: true
+TASKS_DIR: <override>          # e.g., _docs/tasks/  (vs. default _docs/02_tasks/)
+module_layout_path: <override>  # e.g., _docs/tasks/_suite_module_layout.md
+```
+
+When `suite_level: true` is present, the following gate adjustments apply — and ONLY these. All other steps (1–14, 16) execute unchanged:
+
+1. **TASKS_DIR override** is honored throughout the skill (Step 1 Parse, Step 13 Archive, Step 15 input paths if it ran). Default `_docs/02_tasks/` is replaced by the supplied path.
+2. **module_layout_path override** is read instead of the hardcoded `_docs/02_document/module-layout.md` in Step 4 (Assign File Ownership). The supplied file uses the same `Per-Component Mapping` schema. If both the override and the hardcoded path are missing, behavior is unchanged from default mode (STOP and instruct).
+3. **Step 14.5 (Cumulative Code Review) — SKIPPED**. The meta-repo has no `_docs/02_document/architecture_compliance_baseline.md`; cross-task drift is captured by the next `monorepo-status` cycle instead.
+4. **Step 15 (Product Implementation Completeness Gate) — SKIPPED**. The gate's hard inputs (`_docs/02_document/architecture.md`, `system-flows.md`, `components/*/description.md`) do not exist in the meta-repo artifact layout. Suite-level tasks are infrastructure / coordination work (renames, cross-repo edits, suite-root infra additions), not feature implementation; the equivalent completeness signal is the next `monorepo-status` drift report (which the meta-repo flow re-runs immediately after Step 3.5 returns).
+5. **Final report filename**: `_docs/03_implementation/suite_implementation_report_{run_name}.md` (in addition to the existing feature/test/refactor variants). Batch reports follow `_docs/03_implementation/suite_batch_{NN}_report.md`.
+6. **Tracker integration** (Step 5: In Progress, Step 12: In Testing) runs unchanged — suite-level tickets follow the same tracker rules as any other.
+
+Without `suite_level: true`, none of these adjustments apply and the skill runs exactly as documented in default mode.
+
 ## Prerequisite Checks (BLOCKING)
 
-1. `TASKS_DIR/todo/` exists and contains at least one task file — **STOP if missing**
+1. `TASKS_DIR/todo/` exists and contains at least one task file for the selected context — **STOP if missing**
+   - Exception for Product implementation re-entry: if no selected product tasks remain in `todo/`, but the active autodev state is Step 7 or the latest product completeness report is missing/invalid/contains `FAIL`, skip directly to Step 15 (Product Implementation Completeness Gate). This gate may create remediation tasks and return to Step 1. Do not write a final implementation report from this state.
 2. `_dependencies_table.md` exists — **STOP if missing**
 3. At least one task is not yet completed — **STOP if all done**
 4. **Working tree is clean** — run `git status --porcelain`; the output must be empty.
@@ -62,9 +103,9 @@ TASKS_DIR/
 
 ### 1. Parse
 
-- Read all task `*.md` files from `TASKS_DIR/todo/` (excluding files starting with `_`)
+- Read selected task `*.md` files from `TASKS_DIR/todo/` (excluding files starting with `_`)
 - Read `_dependencies_table.md` — parse into a dependency graph (DAG)
-- Validate: no circular dependencies, all referenced dependencies exist
+- Validate: no circular dependencies in the selected task graph, all referenced selected-task dependencies exist or are already completed in `TASKS_DIR/done/`
 
 ### 2. Detect Progress
 
@@ -83,7 +124,7 @@ TASKS_DIR/
 
 ### 4. Assign File Ownership
 
-The authoritative file-ownership map is `_docs/02_document/module-layout.md` (produced by the decompose skill's Step 1.5). Task specs are purely behavioral — they do NOT carry file paths. Derive ownership from the layout, not from the task spec's prose.
+The authoritative file-ownership map is `_docs/02_document/module-layout.md` (produced by the decompose skill's Step 1.5), unless `suite_level: true` was supplied in the invocation context — in which case the `module_layout_path` override is read instead (see "Suite-level invocation context" above). Task specs are purely behavioral — they do NOT carry file paths. Derive ownership from the layout, not from the task spec's prose.
 
 For each task in the batch:
 - Read the task spec's **Component** field.
@@ -102,7 +143,7 @@ If `_docs/02_document/module-layout.md` is missing or the component is not found
 
 ### 5. Update Tracker Status → In Progress
 
-For each task in the batch, transition its ticket status to **In Progress** via the configured work item tracker (see `protocols.md` for tracker detection) before starting work. If `tracker: local`, skip this step.
+For each task in the batch, transition its ticket status to **In Progress** via the configured work item tracker (see `protocols.md` for tracker detection) before starting work. If `tracker: local`, skip this step. If a tracker operation fails unexpectedly, follow `.cursor/rules/tracker.mdc`.
 
 ### 6. Implement Tasks Sequentially
 
@@ -111,7 +152,7 @@ For each task in the batch, transition its ticket status to **In Progress** via
 For each task in the batch **in topological order, one at a time**:
 1. Read the task spec file.
 2. Respect the file-ownership envelope computed in Step 4 (OWNED / READ-ONLY / FORBIDDEN).
-3. Implement the feature and write/update tests for every acceptance criterion in the spec. If a test cannot run in the current environment (e.g., TensorRT requires GPU), the test must still be written and skip with a clear reason.
+3. Implement the feature and write/update tests for every acceptance criterion in the spec. Tests for internal product behavior must exercise the production implementation path. If a test cannot run in the current environment (e.g., TensorRT requires GPU), the test must still exist and skip/block with a clear prerequisite reason, but that skip does not make missing production code complete.
 4. Run the relevant tests locally before moving on to the next task in the batch. If tests fail, fix in-place — do not defer.
 5. Capture a short per-task status line (files changed, tests pass/fail, any blockers) for the batch report.
 
@@ -188,18 +229,22 @@ Track `auto_fix_attempts` and `escalated_findings` in the batch report for retro
 
 ### 12. Update Tracker Status → In Testing
 
-After the batch is committed and pushed, transition the ticket status of each task in the batch to **In Testing** via the configured work item tracker. If `tracker: local`, skip this step.
+After the batch is committed (and pushed if the user approved pushing), transition the ticket status of each task in the batch to **In Testing** via the configured work item tracker. If `tracker: local`, skip this step. If a tracker operation fails unexpectedly, follow `.cursor/rules/tracker.mdc`.
 
 ### 13. Archive Completed Tasks
 
 Move each completed task file from `TASKS_DIR/todo/` to `TASKS_DIR/done/`.
 
+For product implementation, this archive means "batch implementation accepted." The Product Implementation Completeness Gate can still require follow-up remediation tasks before the feature is complete; it does not move original task files back to `todo/`.
+
 ### 14. Loop
 
 - Go back to step 2 until all tasks in `todo/` are done
 
 ### 14.5. Cumulative Code Review (every K batches)
 
+**Skipped entirely when `suite_level: true`** (see "Suite-level invocation context" above) — the meta-repo has no `architecture_compliance_baseline.md` to evaluate against; cross-task drift is captured by the next `monorepo-status` cycle.
+
 - **Trigger**: every K completed batches (default `K = 3`; configurable per run via a `cumulative_review_interval` knob in the invocation context)
 - **Purpose**: per-batch review (Step 9) catches batch-local issues; cumulative review catches issues that only appear when tasks are combined — architecture drift, cross-task inconsistency, duplicate symbols introduced across different batches, contracts that drifted across producer/consumer batches
 - **Scope**: the union of files changed since the **last** cumulative review (or since the start of the run if this is the first)
@@ -215,22 +260,108 @@ Move each completed task file from `TASKS_DIR/todo/` to `TASKS_DIR/done/`.
 - **Interaction with Auto-Fix Gate**: Architecture findings (new category from code-review Phase 7) always escalate per the implement auto-fix matrix; they cannot silently auto-fix
 - **Resumability**: if interrupted, the next invocation checks for the latest `cumulative_review_batches_*.md` and computes the changed-file set from batch reports produced after that review
 
-### 15. Final Test Run
+### 15. Product Implementation Completeness Gate
 
-- After all batches are complete, run the full test suite once
-- Read and execute `.cursor/skills/test-run/SKILL.md` (detect runner, run suite, diagnose failures, present blocking choices)
-- Test failures are a **blocking gate** — do not proceed until the test-run skill completes with a user decision
-- When tests pass, report final summary
+Run this gate after all **product implementation** tasks are complete and before writing any final product implementation report or allowing autodev to proceed to testability/test decomposition. Skip this gate when (a) the remaining context is explicitly test implementation or refactoring (as determined by the task files and report filename rules), OR (b) `suite_level: true` was supplied in the invocation context (the gate's inputs do not exist in the meta-repo artifact layout — see "Suite-level invocation context" above).
+
+**Goal**: catch the failure mode where narrow tests validate scaffold behavior while the task's actual outcome, included scope, architecture promise, or named integration remains unimplemented.
+
+Inputs:
+
+- Completed product task specs from `_docs/02_tasks/done/` for the current cycle
+- `_docs/02_document/architecture.md`
+- `_docs/02_document/system-flows.md`
+- Relevant `_docs/02_document/components/*/description.md` files
+- Current source code under each completed task's ownership envelope
+- Batch reports and code-review reports for the current cycle
+
+For each completed product task:
+
+1. Read these sections from the task spec: `Description`, `Outcome`, `Scope / Included`, `Acceptance Criteria`, `Non-Functional Requirements`, `Constraints`, and explicit named technologies or integrations.
+2. Compare those promises against actual source code, not only tests or report prose.
+3. Search the task's owned component files for unresolved implementation markers: `placeholder`, `stub`, `reserved`, `TODO`, `NotImplemented`, `pass`, `deterministic`, `fake`, `mock`, `scaffold`, `native bridge`, and empty native/readme-only integration directories. Ignore test fixtures/mocks only when they are under test-owned paths and not used as production behavior.
+4. Verify that each named runtime dependency in the task promise is integrated as production behavior, not merely represented by an interface. Examples: if a task promises FAISS, DINOv2, BASALT, LightGlue, OpenCV, RANSAC, a database, cloud service, or hardware SDK, the production code must either call that dependency or contain an adapter that loads and executes the real dependency package. A deterministic fallback, fake runner, empty `native/` package, or "bridge to be supplied later" is **FAIL** unless the task itself explicitly scoped the dependency out before implementation started.
+5. Distinguish internal implementation from external prerequisites:
+   - Internal product capabilities (VIO, anchor verification, cache retrieval, safety wrapper, FDR, MAVLink emission) must be implemented in production code before the task can pass.
+   - External systems/hardware/data (Jetson device, physical camera, ArduPilot process, QGC, third-party service credentials, unavailable licensed dataset) may be `BLOCKED` only when production code exists and the missing prerequisite is outside the product boundary.
+6. Verify tests exercise the real implementation path where local prerequisites exist. Environment-gated tests may skip only with an explicit prerequisite reason; they do not make missing production code complete.
+7. For any architecture promise that describes an end-to-end user outcome, verify there is an executable production pipeline connecting the relevant components. Isolated component contracts and test-only harness orchestration are not enough.
+8. Classify each task:
+   - **PASS**: task promises are implemented or explicitly out of scope in the task itself.
+   - **BLOCKED**: production code exists but cannot be fully verified due to external hardware/data/license/runtime prerequisites; the blocker is explicit and tests report blocked/skipped with reason.
+   - **FAIL**: promised production behavior is missing, only scaffolded, or only represented in tests/reports.
+
+#### 15.b System-Pipeline Check (runs ONCE per gate invocation, after per-task classification)
+
+The per-task classification above (steps 1–8) operates on `_docs/02_tasks/done/`. It catches missing component-local behavior but it CANNOT catch a missing *integration* — there is no task to fail if no task ever owned the integration in the first place. The GPS-passthrough incident (May 2026) escaped this gate because every per-component task in `done/` was honestly complete; the missing piece was the cross-component loop, which had no owning task.
+
+The system-pipeline check fixes that by walking the architecture documents directly, independent of `done/`.
+
+**Inputs**:
+- `_docs/02_document/architecture.md`
+- `_docs/02_document/system-flows.md`
+- Full source tree under the project's production directory (e.g. `src/`).
+
+**Procedure**:
+
+1. **Enumerate end-to-end pipelines.** Read `architecture.md` and `system-flows.md`. For each named pipeline / operational flow that spans 2+ components, record the ordered component sequence and the trigger (per-frame, per-request, scheduled, manual).
+2. **Grep for production callers of each seam method.** For each adjacent pair `A → B` in a pipeline, find a production source file (not under `tests/`, not under a `bench/` package, not a doc) that calls `A`'s public output method AND passes the result into `B`'s public input method.
+3. **Classify the pipeline**:
+   - **WIRED**: a production caller exists and the chain is complete from the first to the last component in the sequence.
+   - **PARTIALLY WIRED**: some adjacent pairs have callers but at least one seam is missing.
+   - **NOT WIRED**: no production code calls the pipeline's components in order. Bench tools, unit tests, and microbenchmarks do NOT count as "wiring".
+4. **Distinguish "wired but stubbed" from "wired with real components"**: a caller that invokes a passthrough / GPS-from-tlog / mock-output-generator instead of the real component is `NOT WIRED` for the purposes of this gate. The seam exists in the source file but the production behavior is faked. Grep for the same scaffold markers Step 15 already enumerates (`placeholder`, `stub`, `passthrough`, `scaffold until`, etc.) inside the caller's body.
+5. **Output**: append a `## System Pipeline Audit` section to `_docs/03_implementation/implementation_completeness_cycle[N]_report.md`. Per-pipeline row: name, sequence, classification, evidence file (the caller, or "NONE FOUND"), remediation suggestion if not `WIRED`.
+
+**Pipeline classification feeds the combined gate below.** Any pipeline that is not `WIRED` is a system-level FAIL that the per-task gate cannot rescue.
+
+**Why this is here and not only in decompose**: decompose Step 1.7 creates integration tasks up front; this check verifies the integration tasks actually got implemented (or, if they were never created, surfaces the gap before the cycle closes). The two layers are belt-and-suspenders by design.
+
+Save the audit to `_docs/03_implementation/implementation_completeness_cycle[N]_report.md` with:
+
+- Per-task classification
+- Evidence files/symbols checked
+- Any unresolved scaffold/native placeholders
+- Any named promised technologies not integrated
+- **System Pipeline Audit table** (per pipeline: name, sequence, WIRED / PARTIALLY WIRED / NOT WIRED, evidence file, remediation suggestion)
+- Required remediation task suggestions, each sized to 5 points or less
+
+Gate:
+
+- If every product task is `PASS` or `BLOCKED` with explicit prerequisite evidence, AND every enumerated pipeline is `WIRED`, continue to Final Test Run.
+- If any product task is `FAIL` OR any pipeline is `PARTIALLY WIRED` / `NOT WIRED`, STOP. Do not write the final product implementation report and do not proceed to any downstream autodev step. Completed original task files remain in `done/`; the missing work is represented by remediation tasks. Present a Choose block:
+  - A) Create remediation tasks now and return to implementation. (For pipeline FAILs the remediation task is a NEW integration task owned by the spine component per `_docs/02_document/module-layout.md`; it is NOT a test task and NOT a doc task; its deliverable is production code that drives the pipeline against real components.)
+  - B) Mark the missing behavior explicitly out of scope in task/docs, then re-run this gate
+  - C) Abort for manual correction
+- Recommendation must normally be A unless the user deliberately accepts reduced scope.
+
+Remediation task creation:
+
+1. For each `FAIL`, create one or more task specs using `.cursor/skills/decompose/templates/task.md`; each remediation task must be sized at 5 points or less.
+2. Save each task to `_docs/02_tasks/todo/` with a short name prefixed by `remediate_`.
+3. Set **Component** to the failed task's component and set **Dependencies** to the failed task ID plus any remediation prerequisites.
+4. Create or defer tracker tickets using the same tracker rules as decompose/new-task: if tracker is available, create tickets immediately; if the user explicitly chose `tracker: local`, keep numeric prefixes with `Tracker: pending` / `Epic: pending`.
+5. Append the remediation tasks to `_docs/02_tasks/_dependencies_table.md`.
+6. Return to Step 1 (Parse) in **Product implementation** context. The final product implementation report can be written only after remediation tasks complete and this gate reruns without `FAIL`.
+
+### 16. Final Test Run
+
+- After all batches are complete, run the full test suite once unless the invoking flow's immediate next step is `Run Tests`.
+- If the next flow step is `Run Tests`, record a handoff in the final implementation report and let `.cursor/skills/test-run/SKILL.md` own the full-suite gate to avoid duplicate full runs.
+- When this step does run, read and execute `.cursor/skills/test-run/SKILL.md` (detect runner, run suite, diagnose failures, present blocking choices).
+- Test failures are a **blocking gate** — do not proceed until the test-run skill completes with a user decision.
+- When tests pass, report final summary.
 
 ## Batch Report Persistence
 
-After each batch completes, save the batch report to `_docs/03_implementation/batch_[NN]_cycle[N]_report.md` for feature implementation (or `batch_[NN]_report.md` for test/refactor runs). Create the directory if it doesn't exist. When all tasks are complete, produce a FINAL implementation report with a summary of all batches. The filename depends on context:
+After each batch completes, save the batch report to `_docs/03_implementation/batch_[NN]_cycle[N]_report.md` for feature implementation (or `batch_[NN]_report.md` for test/refactor runs). Create the directory if it doesn't exist. For product implementation, produce the FINAL implementation report only after the Product Implementation Completeness Gate passes. For test and refactor implementation, produce the FINAL report after all selected tasks complete and the full-suite gate is either run or handed off per Step 16. The filename depends on context:
 
 - **Test implementation** (tasks from test decomposition): `_docs/03_implementation/implementation_report_tests.md`
 - **Feature implementation**: `_docs/03_implementation/implementation_report_{feature_slug}_cycle{N}.md` where `{feature_slug}` is derived from the batch task names (e.g., `implementation_report_core_api_cycle2.md`) and `{N}` is the current `state.cycle` from `_docs/_autodev_state.md`. If `state.cycle` is absent (pre-migration), default to `cycle1`.
 - **Refactoring**: `_docs/03_implementation/implementation_report_refactor_{run_name}.md`
+- **Suite-level** (when `suite_level: true` was supplied — see "Suite-level invocation context" above): `_docs/03_implementation/suite_implementation_report_{run_name}.md`. Batch reports use `_docs/03_implementation/suite_batch_{NN}_report.md`. `{run_name}` is derived from the batch task IDs (e.g., `suite_implementation_report_az543_az549_az550.md`).
 
-Determine the context from the task files being implemented: if all tasks have test-related names or belong to a test epic, use the tests filename; otherwise derive the feature slug from the component names and append the cycle suffix.
+Determine the context from the task files being implemented: if all tasks have test-related names or belong to a test epic, use the tests filename; if `suite_level: true` was supplied, use the suite filename; otherwise derive the feature slug from the component names and append the cycle suffix.
 
 Batch report filenames must also include the cycle counter when running feature implementation: `_docs/03_implementation/batch_{NN}_cycle{N}_report.md` (test and refactor runs may use the plain `batch_{NN}_report.md` form since they are not cycle-scoped).
 
@@ -266,6 +397,7 @@ After each batch, produce a structured report:
 | Same task rewritten 3+ times without green tests | Mark Blocked, continue batch, escalate at batch end |
 | Task blocked on external dependency (not in task list) | Report and skip |
 | File ownership violated (task wrote outside OWNED) | ASK user |
+| Product completeness gate finds missing promised implementation | STOP — create remediation tasks or get explicit user scope reduction |
 | Test failure after final test run | Delegate to test-run skill — blocking gate |
 | All tasks complete | Report final summary, suggest final commit |
 | `_dependencies_table.md` missing | STOP — run `/decompose` first |
@@ -283,4 +415,5 @@ Each batch commit serves as a rollback checkpoint. If recovery is needed:
 - Never start a task whose dependencies are not yet completed
 - Never run tasks in parallel and never spawn subagents — see `.cursor/rules/no-subagents.mdc`
 - If a task is flagged as stuck, stop working on it and report — do not let it loop indefinitely
-- Always run the full test suite after all batches complete (step 15)
+- Always run the Product Implementation Completeness Gate before final product reports
+- Always run or hand off the full test suite after all batches complete (step 16)

.cursor/skills/new-task/SKILL.md

@@ -84,29 +84,66 @@ Assess the change along these dimensions:
 - **Novelty**: does it involve libraries, protocols, or patterns not already in the codebase?
 - **Risk**: could it break existing functionality or require architectural changes?
 
-Classification:
+### 2a. Complexity-Points Estimate
+
+Project policy (per the workspace user-rule on ADO points): aim for tasks at 2–3 points (rarely 5). Tasks at 8 points are high risk; tasks at 13 are too complex and MUST be broken down. The new-task skill enforces this here, before producing a single-file task spec.
+
+Map the Scope/Novelty/Risk profile to a points estimate using this table:
+
+| Profile | Points | Examples |
+|---------|--------|----------|
+| All three low | **1–2** | One-line config change; trivial CRUD field addition |
+| Two low + one medium | **3** | Localized refactor; add one well-understood endpoint |
+| One low + two medium, OR all medium | **5** | New small feature touching 2–3 components; integration with a known library |
+| Any high, OR two medium + one high | **8** | Cross-cutting concern across 4+ components; integration with an unfamiliar protocol; significant architectural change |
+| Two or three high | **13** | New subsystem; unfamiliar tech across the stack; multiple unknown unknowns |
+
+If a relevant LESSONS.md entry biases the estimate (e.g., "auth-related changes historically take 2× estimate"), apply the multiplier and round up to the next discrete point on the scale (1, 2, 3, 5, 8, 13).
+
+### 2b. Routing by Complexity
+
+| Estimate | Default routing | Override path |
+|----------|-----------------|---------------|
+| **1–5** | Continue this skill at Step 3 (Research) or Step 4 (Codebase Analysis) — see classification below | — |
+| **8** | **STOP this skill and recommend handoff to `/decompose @<feature_description>`** (single-component decompose mode if the affected scope fits inside one component, default mode if it does not). The user may override and proceed in `/new-task`, but the override must be explicitly chosen. | C) Proceed in /new-task anyway with the user's acknowledgement that the resulting task is high-risk and may need to be re-decomposed mid-implementation |
+| **13** | **STOP this skill — auto-handoff is mandatory.** A 13-point feature cannot be a single task spec. Invoke `/decompose @<feature_description>` (default mode) before writing any task file. Surface the handoff to the user with no override path; this is a hard policy gate. | None — must decompose |
+
+For the auto-handoff path:
+
+1. Render a one-paragraph description of the feature suitable to feed `/decompose` (combine Step 1's verbatim user description with the complexity-points reasoning).
+2. Save it to `_docs/02_task_plans/<feature_slug>/feature-description.md` so the decompose skill has a stable input file.
+3. Either (a) directly auto-chain into `.cursor/skills/decompose/SKILL.md` in default mode with this file as input, or (b) report the handoff to the user along with the exact `/decompose` invocation and stop. Pick (a) only if the user has explicitly enabled auto-chain across skills (e.g., we are inside an `/autodev` invocation); otherwise pick (b).
+
+### 2c. Research vs Skip Research (only for ≤5 estimates)
+
+Classification (independent of points; runs only when points ≤ 5 and Step 2b chose Continue):
 
 | Category | Criteria | Action |
 |----------|----------|--------|
-| **Needs research** | New libraries/frameworks, unfamiliar protocols, significant architectural change, multiple unknowns | Proceed to Step 3 (Research) |
+| **Needs research** | New libraries/frameworks, unfamiliar protocols, multiple unknowns | Proceed to Step 3 (Research) |
 | **Skip research** | Extends existing functionality, uses patterns already in codebase, straightforward new component with known tech | Skip to Step 4 (Codebase Analysis) |
 
-Present the assessment to the user:
+Present the full assessment to the user:
 
 ```
 ══════════════════════════════════════
  COMPLEXITY ASSESSMENT
 ══════════════════════════════════════
- Scope:   [low / medium / high]
- Novelty: [low / medium / high]
- Risk:    [low / medium / high]
+ Scope:    [low / medium / high]
+ Novelty:  [low / medium / high]
+ Risk:     [low / medium / high]
+ Points:   [1 / 2 / 3 / 5 / 8 / 13]   (project aim: 2–3, rarely 5)
+ Routing:  [Continue in /new-task | Hand off to /decompose]
 ══════════════════════════════════════
- Recommendation: [Research needed / Skip research]
- Reason: [one-line justification]
+ Recommendation: [Research needed | Skip research | Decompose required]
+ Reason: [one-line justification, including any LESSONS.md influence]
 ══════════════════════════════════════
 ```
 
-**BLOCKING**: Ask the user to confirm or override the recommendation before proceeding.
+**BLOCKING**:
+- If points ≤ 5 → ask the user to confirm or override the research recommendation before proceeding.
+- If points = 8 → ask the user to choose between hand-off to /decompose (recommended) and continuing in /new-task with explicit risk acknowledgement.
+- If points = 13 → STOP and present the handoff plan; do not offer a continue-anyway override.
 
 ---
 
@@ -203,7 +240,13 @@ Apply the four shared-task triggers from `.cursor/skills/decompose/SKILL.md` Ste
   2. Add the layout edit to the task's deliverables; the implementer writes it alongside the code change.
   3. If `module-layout.md` does not exist, STOP and instruct the user to run `/document` first (existing-code flow) or `/decompose` default mode (greenfield). Do not guess.
 
-Record the classification and any contract/layout deliverables in the working notes; they feed Step 5 (Validate Assumptions) and Step 6 (Create Task).
+- **ADR cross-check** — runs unconditionally for every new-task in any of the three classifications above:
+  1. If `_docs/02_document/adr/` exists, scan every `Status: Accepted` ADR. For each, ask: "would the proposed task either contradict this ADR's `Decision` or materially affect its `Consequences`?"
+  2. **Conflict** (task contradicts an Accepted ADR) → STOP and Choose A/B/C: **A)** Re-scope the task to comply with the ADR, **B)** Propose superseding the ADR — the task spec then includes a deliverable to invoke `/plan --adr-only` (or the next `/plan` cycle's Step 4.5) with `Supersedes: ADR-NNN`, and the new task does NOT proceed until that supersede ADR is `Accepted`, **C)** Park the task in `backlog/` with a `Blocked-By: ADR-NNN review` note. Do not silently approve a contradictory task.
+  3. **Drift** (task changes assumptions an ADR depends on but does not directly contradict it) → record the affected ADR(s) under a new `### ADR Impact` section in the task spec with `> Affects ADR NNN_<slug>: <one-line summary>`. The implementer surfaces this at code-review Phase 7 (which then classifies it as ADR-Drift if not addressed).
+  4. **Aligned** (task implements something an Accepted ADR mandates) → cite the ADR(s) under `### ADR Compliance` in the task spec with `> Implements ADR NNN_<slug>`. Code-review Phase 7 then expects matching evidence in the implemented code.
+
+Record the classification, any contract/layout deliverables, and any ADR cross-check outcomes in the working notes; they feed Step 5 (Validate Assumptions) and Step 6 (Create Task).
 
 **BLOCKING**: none — this step surfaces findings; the user confirms them in Step 5.
 
@@ -263,6 +306,9 @@ Present using the Choose format for each decision that has meaningful alternativ
 - [ ] If Step 4.5 classified the task as producer, the `## Contract` section exists and points at a contract file
 - [ ] If Step 4.5 classified the task as consumer, `### Document Dependencies` lists the relevant contract file
 - [ ] If Step 4.5 flagged a layout delta, the task's Scope.Included names the `module-layout.md` edit
+- [ ] If Step 4.5 flagged an ADR conflict, the task is either re-scoped (A), explicitly blocked on a supersede ADR (B), or parked in backlog (C) — never silently bypassed
+- [ ] If Step 4.5 flagged ADR drift, the task spec has an `### ADR Impact` section listing the affected ADR(s)
+- [ ] If Step 4.5 flagged ADR alignment, the task spec has an `### ADR Compliance` section citing the implemented ADR(s)
 
 ---
 
@@ -282,7 +328,7 @@ Present using the Choose format for each decision that has meaningful alternativ
    - Update **Epic** field: `[EPIC-ID]`
 3. Rename the file from `[##]_[short_name].md` to `[TICKET-ID]_[short_name].md`
 
-If the work item tracker is not authenticated or unavailable (`tracker: local`):
+If the work item tracker is not authenticated or unavailable, follow `.cursor/rules/tracker.mdc` before continuing. Only if the user explicitly chooses `tracker: local`:
 - Keep the numeric prefix
 - Set **Tracker** to `pending`
 - Set **Epic** to `pending`
@@ -337,7 +383,7 @@ After the user chooses **Done**:
 | Research skill hits a blocker | Follow research skill's own escalation rules |
 | Codebase analysis reveals conflicting architectures | **ASK** user which pattern to follow |
 | Complexity exceeds 5 points | **WARN** user and suggest splitting into multiple tasks |
-| Work item tracker MCP unavailable | **WARN**, continue with local-only task files |
+| Work item tracker MCP unavailable | Follow `.cursor/rules/tracker.mdc`; do not continue in local mode unless the user explicitly chooses it |
 
 ## Trigger Conditions
 

.cursor/skills/plan/SKILL.md

@@ -15,7 +15,7 @@ disable-model-invocation: true
 
 # Solution Planning
 
-Decompose a problem and solution into architecture, data model, deployment plan, system flows, components, tests, and work item epics through a systematic 6-step workflow.
+Decompose a problem and solution into architecture, data model, deployment plan, system flows, components, ADRs, tests, and work item epics through a systematic workflow with seven step files (1, 2, 3, 4, 4.5, 5, 6) plus a Final quality checklist.
 
 ## Core Principles
 
@@ -55,7 +55,7 @@ Read `steps/01_artifact-management.md` for directory structure, save timing, sav
 
 ## Progress Tracking
 
-At the start of execution, create a TodoWrite with all steps (1 through 6 plus Final). Update status as each step completes.
+At the start of execution, create a TodoWrite with all steps (1, 2, 3, 4, 4.5, 5, 6 plus Final). Update status as each step completes. The fractional Step 4.5 (ADR Capture) sits between Architecture Review (Step 4) and Test Specifications (Step 5).
 
 ## Workflow
 
@@ -85,6 +85,16 @@ Read and follow `steps/04_review-risk.md`.
 
 ---
 
+### Step 4.5: Architecture Decision Records (ADRs)
+
+Read and follow `steps/04-5_adr-capture.md`.
+
+This step captures the architecture and tech-stack decisions that were made (or revised) in Steps 2–4 as durable, dated, immutable records under `_docs/02_document/adr/`. ADRs are the single thing in `_docs/` that explain the **why** of each major decision after the conversation history is gone. They are consumed by `decompose` (when bootstrapping module layout), `new-task` (when assessing a new feature against existing decisions), `refactor` (when proposing replacements), and any future code-review cycle that needs to confirm a structural choice was deliberate.
+
+This step is **BLOCKING**: the ADR set must be reviewed and confirmed by the user before Step 5 begins.
+
+---
+
 ### Step 5: Test Specifications
 
 Read and follow `steps/05_test-specifications.md`.
@@ -120,7 +130,7 @@ Read and follow `steps/07_quality-checklist.md`.
 |-----------|--------|
 | Missing acceptance_criteria.md, restrictions.md, or input_data/ | **STOP** — planning cannot proceed |
 | Ambiguous requirements | ASK user |
-| Input data coverage below 75% | Search internet for supplementary data, ASK user to validate |
+| Input data coverage below the canonical threshold (`cursor-meta.mdc` Quality Thresholds) | Search internet for supplementary data, ASK user to validate |
 | Technology choice with multiple valid options | ASK user |
 | Component naming | PROCEED, confirm at next BLOCKING gate |
 | File structure within templates | PROCEED |
@@ -146,6 +156,8 @@ Read and follow `steps/07_quality-checklist.md`.
 │    [BLOCKING: user confirms components]                        │
 │ 4. Review & Risk       → risk register, iterations              │
 │    [BLOCKING: user confirms mitigations]                       │
+│ 4.5 ADR Capture        → _docs/02_document/adr/NNN_*.md         │
+│    [BLOCKING: user confirms ADR set]                           │
 │ 5. Test Specifications → per-component test specs               │
 │ 6. Work Item Epics     → epic per component + bootstrap         │
 │    ─────────────────────────────────────────────────           │

.cursor/skills/plan/steps/01_artifact-management.md

@@ -26,6 +26,10 @@ DOCUMENT_DIR/
 │   └── deployment_procedures.md
 ├── risk_mitigations.md
 ├── risk_mitigations_02.md          (iterative, ## as sequence)
+├── adr/
+│   ├── 001_[decision_slug].md
+│   ├── 002_[decision_slug].md
+│   └── ...
 ├── components/
 │   ├── 01_[name]/
 │   │   ├── description.md
@@ -66,6 +70,8 @@ DOCUMENT_DIR/
 | Step 3 | Common helpers generated | `common-helpers/[##]_helper_[name].md` |
 | Step 3 | Diagrams generated | `diagrams/` |
 | Step 4 | Risk assessment complete | `risk_mitigations.md` |
+| Step 4.5 | Each ADR captured | `adr/NNN_[decision_slug].md` |
+| Step 4.5 | ADR index updated | `adr/README.md` |
 | Step 5 | Tests written per component | `components/[##]_[name]/tests.md` |
 | Step 6 | Epics created in work item tracker | Tracker via MCP |
 | Final | All steps complete | `FINAL_report.md` |
@@ -85,3 +91,15 @@ If DOCUMENT_DIR already contains artifacts:
 2. Identify the last completed step based on which artifacts exist
 3. Resume from the next incomplete step
 4. Inform the user which steps are being skipped
+
+#### Step 4.5 (ADR Capture) resumption rule
+
+ADR files have a `Status` field that disambiguates "step in progress" from "step done":
+
+- `Status: Proposed` → Step 4.5 is **in progress**. The user has not yet hit the BLOCKING gate (or hit it and chose B/C/D, which kept files at `Proposed`). Resume Step 4.5 at Phase 4.5f and re-present the BLOCKING Choose to the user. Do NOT skip to Step 5.
+- `Status: Accepted` AND `adr/README.md` index exists AND every Accepted ADR is referenced in the index → Step 4.5 is **done**. Skip to Step 5.
+- `Status: Accepted` but `adr/README.md` is missing or out of date → Step 4.5 is **partially complete**. Resume at Phase 4.5d (Maintain the ADR Index) before moving on.
+- Mixed `Proposed` + `Accepted` files in the same directory → Step 4.5 is **in progress** with prior partial confirmations. Resume at Phase 4.5f and re-present only the still-`Proposed` ADRs.
+- Empty `adr/` directory or no `adr/` directory → Step 4.5 has not started yet. Begin at Phase 4.5a.
+
+The `Date` field on every Accepted ADR is the date the user confirmed it; do not regenerate it during resumption.

.cursor/skills/plan/steps/04-5_adr-capture.md

@@ -0,0 +1,187 @@
+# Step 4.5: Architecture Decision Records (ADRs)
+
+**Role**: Architect / technical writer
+**Goal**: Capture every major architecture, tech-stack, data-model, and integration decision made during Steps 2–4 as a durable, dated, immutable record under `_docs/02_document/adr/`.
+**Constraints**: ADRs only — do not re-open architecture; do not make new decisions in this step. Document what has been decided, not what is still open.
+
+ADRs are the single thing in `_docs/` that explains the **why** of each major decision after the conversation history is gone. They are consumed by:
+
+- `decompose` Step 1.5 (`steps/01-5_module-layout.md`) — every Accepted ADR is cross-checked against the module-layout proposal; conflicts trigger an explicit Choose between supersede / exception / re-open.
+- `new-task` Step 4.5 (`SKILL.md` § "Step 4.5: Contract & Layout Check") — every new task is classified against Accepted ADRs as Conflict / Drift / Aligned; conflicts STOP the task with a Choose A/B/C; drift adds an `### ADR Impact` section; alignment adds an `### ADR Compliance` section.
+- `refactor` Phase 2b.1 (`phases/02-analysis.md`) — every Accepted ADR is diffed against the proposed roadmap; Violations trigger a BLOCKING supersede gate that produces a `supersede_adr_NNN.md` task before any refactor task is created.
+- `code-review` Phase 7 (`SKILL.md` § "Phase 7: Architecture Compliance") — every changed-files batch is checked against Accepted ADRs; ADR-Violation findings are Critical, ADR-Drift findings are High.
+
+Discipline that still relies on the human: when a downstream skill detects a Drift case, the resulting task spec MUST land its `## ADR Impact` / `## ADR Compliance` section; the implementer must address it; the next code-review batch then has the context it needs. Drift left undocumented is the silent-failure path — every consumer hook above is designed to make it visible.
+
+## Inputs
+
+- `_docs/02_document/architecture.md` (incl. confirmed `## Architecture Vision`)
+- `_docs/02_document/glossary.md`
+- `_docs/02_document/data_model.md`
+- `_docs/02_document/system-flows.md`
+- `_docs/02_document/risk_mitigations.md` (and any `risk_mitigations_NN.md` iterations from Step 4)
+- `_docs/02_document/components/[##]_[name]/description.md`
+- `_docs/02_document/deployment/` (CI/CD, environments, observability)
+- `_docs/00_problem/restrictions.md` and `_docs/00_problem/acceptance_criteria.md` (each ADR must reference relevant constraints / AC by ID)
+- Optional: `_docs/01_solution/solution.md` and `_docs/01_solution/tech_stack.md` (research output)
+- Optional: `_docs/LESSONS.md` — surface any lesson categories of `architecture` / `dependencies` that bias the recommendation
+
+## What is an ADR (and what is not)
+
+Capture an ADR when **all** of the following hold:
+
+1. The decision picks between two or more genuinely valid approaches with meaningful trade-offs.
+2. The decision has **downstream consequences** that other decisions, code, or tasks inherit from.
+3. The decision is **non-obvious** to a future reader who only sees the final code — they would ask "why was it built this way?" rather than discovering the answer by reading the source.
+
+Do NOT create an ADR for:
+
+- Naming, formatting, or purely cosmetic choices.
+- A choice that is fully implied by a single explicit restriction (`restrictions.md` is itself the record — link to it from the architecture doc instead).
+- A choice the team has not actually made yet — open questions live in `risk_mitigations.md` or `_docs/_process_leftovers/`, not in ADRs.
+- A technology selection where research already produced an exact-fit selection with one viable option (the research doc is the record — link to the relevant `solution_draft*.md` section).
+
+## Process
+
+### Phase 4.5a: Decision Inventory
+
+Walk the inputs and list candidate decisions. For each candidate, record a one-liner:
+
+```
+- [decision] — [trade-off summary] — [downstream consumers] — [evidence file:section]
+```
+
+Inspect at minimum:
+
+| Inspection target | Typical decisions surfaced |
+|-------------------|----------------------------|
+| `architecture.md` § layering | Layering style (clean vs hex vs n-tier), which layer owns transactions, how cross-cutting concerns enter |
+| `architecture.md` § Architecture Vision | The North Star principle (e.g., "edge-first, sync-second"); ADR captures the implication for one specific subsystem |
+| `data_model.md` | Datastore choice (Postgres vs Mongo), partitioning, soft vs hard deletes, schema evolution strategy |
+| `system-flows.md` | Sync vs async boundaries, idempotency strategy, retry policy ownership, error envelope shape |
+| `components/*/description.md` § interfaces | Public-API style (REST vs RPC vs event), versioning strategy, auth/authorization placement |
+| `deployment/containerization.md` | Single container vs sidecar vs init container, base image lineage |
+| `deployment/ci_cd_pipeline.md` | Trunk-based vs feature-branch, gate ordering, deploy strategy (blue-green / canary / all-at-once) |
+| `deployment/observability.md` | Logging stack, metric backend, sampling rate decisions, retention |
+| `risk_mitigations.md` | Risk-acceptance trade-offs (e.g., "we accept N% data loss in exchange for sub-100ms p99") |
+| Tech-stack from `_docs/01_solution/tech_stack.md` | Anything where research recorded ≥2 candidates and a winner |
+
+Drop any candidate that fails the three "what is an ADR" criteria above. Keep the rest.
+
+### Phase 4.5b: Numbering and Slugs
+
+ADRs are numbered globally per project, monotonically, never re-used.
+
+1. List existing files under `_docs/02_document/adr/` matching `^[0-9]{3}_.+\.md$`.
+2. The next ADR number is `max(existing) + 1`, zero-padded to 3 digits.
+3. The slug is kebab-case, ≤6 words, derived from the decision summary. Example: `001_use-postgres-for-transactional-data.md`, `004_event-driven-cross-component-comms.md`.
+
+### Phase 4.5c: Render One ADR Per Decision
+
+For each kept candidate, render the ADR using `templates/adr.md`. Required sections (do NOT omit any):
+
+| Section | Content |
+|---------|---------|
+| **Number** | `NNN` |
+| **Title** | One-line decision statement (matches slug) |
+| **Status** | `Proposed` (only during Step 4.5 iteration) → `Accepted` (after user confirmation at the BLOCKING gate) |
+| **Date** | YYYY-MM-DD (the date the user confirmed) |
+| **Deciders** | The user (project owner) — the AI is not a decider |
+| **Context** | The problem this decision addresses, including links to AC IDs, restriction IDs, risks, and (where relevant) the research draft section |
+| **Decision** | The chosen approach in one sentence, then the supporting detail |
+| **Alternatives Considered** | Each alternative with a one-line "rejected because…" |
+| **Consequences** | Positive (what becomes easier / cheaper / faster) and negative (what becomes harder / locked in / costly to undo). Be honest — every decision has a downside. |
+| **Supersedes / Superseded by** | Empty initially; updated when a future ADR overturns this one |
+| **Evidence** | File-and-section pointers into `_docs/` showing where the decision is reflected (architecture.md § layering, components/02_*/description.md § interface, etc.) |
+
+After rendering, write each file to `_docs/02_document/adr/NNN_<slug>.md`. Keep `Status: Proposed` until the BLOCKING gate.
+
+### Phase 4.5d: Maintain the ADR Index
+
+Write or update `_docs/02_document/adr/README.md` with this exact shape:
+
+```markdown
+# Architecture Decision Records
+
+This index lists every ADR for this project, in number order. ADRs are immutable once `Accepted` —
+new decisions that overturn a prior ADR are recorded as new ADRs whose `Supersedes` field points
+back, and the original ADR's `Superseded by` field is updated.
+
+| # | Title | Status | Date | Supersedes |
+|---|-------|--------|------|------------|
+| 001 | Use Postgres for transactional data | Accepted | 2026-05-21 | — |
+| 002 | Event-driven cross-component comms | Accepted | 2026-05-21 | — |
+| ... | ... | ... | ... | ... |
+```
+
+Sort by `#` ascending. Include all ADRs ever written, even superseded ones — the audit trail is the point.
+
+### Phase 4.5e: Cross-Link from architecture.md
+
+In `architecture.md`, every section that reflects an ADR decision gets a one-line trailing reference:
+
+```markdown
+> See ADR 001 (Use Postgres for transactional data), ADR 003 (Event-driven cross-component comms).
+```
+
+Place the reference at the end of the section, after the prose. This lets a future reader of `architecture.md` jump straight to the rationale.
+
+### Phase 4.5f: BLOCKING Gate — User Confirmation
+
+Present the ADR set to the user using the Choose format from `.cursor/skills/autodev/protocols.md` (or plain text if AskQuestion is unavailable):
+
+```
+══════════════════════════════════════
+ DECISION REQUIRED: ADR set captured (N records)
+══════════════════════════════════════
+ 001 — [title]
+ 002 — [title]
+ ...
+══════════════════════════════════════
+ A) Accept all ADRs as written
+ B) Edit specific ADRs (numbers and edits)
+ C) Add a missed decision (description)
+ D) Remove an ADR (number and reason)
+══════════════════════════════════════
+ Recommendation: A — review the rendered set and confirm; corrections are quick on Round 2
+══════════════════════════════════════
+```
+
+Loop:
+
+- **A** → flip every ADR's `Status` from `Proposed` to `Accepted`, set `Date` to today's date, save, exit step.
+- **B** → apply edits, re-present the modified ADRs, loop.
+- **C** → run Phase 4.5a–4.5e for the missed decision only, append to the set, re-present, loop.
+- **D** → confirm with the user that the candidate fails the three "what is an ADR" criteria, remove the file, update the index, loop.
+
+Do NOT mark `Accepted` without an explicit user A.
+
+## Self-verification
+
+- [ ] Every kept candidate from Phase 4.5a has a corresponding file under `adr/`
+- [ ] Every ADR has all required sections (none empty except `Supersedes` / `Superseded by`)
+- [ ] `Decision` sections are one-sentence-then-detail, not "we'll figure it out"
+- [ ] `Alternatives Considered` lists at least one rejected alternative per ADR
+- [ ] `Consequences` lists both positive AND negative consequences (an ADR with no negatives is suspect)
+- [ ] `Evidence` points at real `_docs/` sections that exist on disk
+- [ ] `adr/README.md` index lists every file in the directory and matches their `Status` / `Date`
+- [ ] `architecture.md` has a trailing `See ADR …` reference at every section that an ADR reflects
+- [ ] The user confirmed the set via Choose A; every ADR is `Accepted` with today's date
+
+## Common mistakes
+
+- **Re-opening architecture**: Step 4.5 records, it does not decide. If a candidate decision turns out to be unsettled, that's a Step 2 / Step 4 gap — return there, do not paper over it with a wishy-washy ADR.
+- **Decision-of-the-week**: do not write an ADR for every minor pattern choice. The bar is "non-obvious to a future reader". 5–15 ADRs is typical for a planning round; 40+ is over-capture.
+- **Negative consequences left empty**: every real decision has costs. If you cannot name one, the decision was not actually weighed.
+- **Vague evidence**: `architecture.md` is not enough — point at the specific section. `architecture.md § Layering` ≠ `architecture.md`.
+- **Numbering reuse**: never recycle a number from a deleted ADR. The audit trail is more important than tidy numbering.
+- **Superseding without recording**: when a later cycle overturns an ADR, the new ADR must point at the old one via `Supersedes`, AND the old ADR's `Superseded by` field must be updated. Index reflects both. (This is enforced when `decompose` or `refactor` later updates ADRs.)
+
+## Escalation
+
+| Situation | Action |
+|-----------|--------|
+| Candidate decision is unsettled (the team has not actually decided) | Return to the originating step (2 / 3 / 4); do NOT write a placeholder ADR |
+| Two candidates in Phase 4.5a turn out to be the same decision phrased differently | Merge into one ADR, list both phrasings in `Context` |
+| User picks D (remove an ADR) and the AI judges the decision is genuinely worth recording | Surface the disagreement, ASK why the user wants it removed, defer to user |
+| Existing `adr/` directory has files but `adr/README.md` is missing or stale | Rebuild the index from the directory before adding new ADRs |

.cursor/skills/plan/steps/05_test-specifications.md

@@ -2,7 +2,7 @@
 
 **Role**: Professional Quality Assurance Engineer
 
-**Goal**: Write test specs for each component achieving minimum 75% acceptance criteria coverage
+**Goal**: Write test specs for each component achieving the canonical minimum acceptance-criteria coverage (currently 75% — see `.cursor/rules/cursor-meta.mdc` Quality Thresholds; do not restate a different number here)
 
 **Constraints**: Test specs only — no test code. Each test must trace to an acceptance criterion.
 

.cursor/skills/plan/steps/06_work-item-epics.md

@@ -58,4 +58,4 @@ Do NOT create minimal epics with just a summary and short description. The epic
 
 8. **Create "Blackbox Tests" epic** — this epic will parent the blackbox test tasks created by the `/decompose` skill. It covers implementing the test scenarios defined in `tests/`.
 
-**Save action**: Epics created via the configured tracker MCP. Also saved locally in `epics.md` with ticket IDs. If `tracker: local`, save locally only.
+**Save action**: Epics created via the configured tracker MCP. Also saved locally in `epics.md` with ticket IDs. If tracker availability fails, follow `.cursor/rules/tracker.mdc`; only if the user explicitly chooses `tracker: local`, save locally only with pending tracker markers.

.cursor/skills/plan/templates/adr.md

@@ -0,0 +1,67 @@
+# ADR-{NNN}: {decision-title}
+
+- **Status**: {Proposed | Accepted | Deprecated | Superseded}
+- **Date**: {YYYY-MM-DD}
+- **Deciders**: {user / project owner}
+- **Supersedes**: {ADR-NNN | —}
+- **Superseded by**: {ADR-NNN | —}
+
+## Context
+
+What problem does this decision address? Cite the relevant constraint(s), acceptance criterion / criteria, and risk(s) by ID.
+
+- Acceptance criteria addressed: AC-{ID-1}, AC-{ID-2}
+- Restrictions addressed: R-{ID-1}, R-{ID-2}
+- Risks addressed: RISK-{ID-1}
+- Research source (if any): `_docs/01_solution/solution_draftN.md` § {section}
+
+A short paragraph (3–6 sentences) explaining why a choice is required now and what makes it non-trivial. Do not pre-announce the decision here — that goes in `Decision`. Focus on the forces at play (load, scale, team familiarity, hardware constraints, regulatory drivers, third-party limits).
+
+## Decision
+
+One declarative sentence: **"We will …"** Then 1–3 paragraphs of supporting detail explaining how the decision will be implemented at the boundaries between components.
+
+Be specific. "We will use Postgres" is too thin; "We will use Postgres 16 with logical replication for read scaling, restricting JSONB columns to top-level metadata only, with all transactional data in normalized tables" is the right resolution.
+
+## Alternatives Considered
+
+| Alternative | Rejected because |
+|-------------|------------------|
+| {Alt 1 — short label} | {one line: the cost / mismatch / risk that ruled it out, ideally referencing a measurable criterion} |
+| {Alt 2 — short label} | {one line} |
+| {Alt 3 — short label} | {one line} |
+
+At least one rejected alternative is mandatory. If only one option was ever considered, this is not an ADR — link to the source restriction or research selection from the parent doc instead.
+
+## Consequences
+
+### Positive
+
+- {What becomes easier / cheaper / faster, with concrete examples where possible}
+- {…}
+
+### Negative
+
+- {What becomes harder / locked in / costly to undo}
+- {…}
+
+Every real decision has both. If the negatives section is hard to fill, the alternatives were probably not weighed seriously — return to the prior step.
+
+### Neutral / Open
+
+- {What is unchanged but worth flagging for future readers (e.g., "this does not change the auth boundary; auth remains in component 02_user_management as decided in ADR-003")}
+
+## Evidence
+
+Where this decision is reflected on disk. Use `file:section` links so future readers can jump.
+
+- `_docs/02_document/architecture.md` § {section}
+- `_docs/02_document/data_model.md` § {section}
+- `_docs/02_document/components/{##_name}/description.md` § {section}
+- `_docs/02_document/system-flows.md` § {flow name}
+- `_docs/02_document/deployment/{file}.md` § {section}
+- {add more as needed}
+
+## Notes
+
+Optional. Use for caveats that did not fit above, links to external research, or follow-ups that the team agreed to revisit on a known trigger ("re-evaluate after 6 months in production" / "re-evaluate when load exceeds 10× baseline").

.cursor/skills/plan/templates/epic-spec.md

@@ -133,4 +133,4 @@ Link to architecture.md and relevant component spec.]
   - `component` — a normal per-component epic
   - `cross-cutting` — a shared concern that spans ≥2 components
   - `tests` — the blackbox-tests epic (always exactly one)
-- Complexity points for child issues follow the project standard: 1, 2, 3, 5, 8. Do not create issues above 5 points — split them.
+- Complexity points for child issues follow the project standard: 1, 2, 3, 5. Do not create issues above 5 points — split them.

.cursor/skills/plan/templates/final-report.md

@@ -1,6 +1,6 @@
 # Final Planning Report Template
 
-Use this template after completing all 6 steps and the quality checklist. Save as `_docs/02_document/FINAL_report.md`.
+Use this template after completing all steps (1, 2, 3, 4, 4.5, 5, 6) and the quality checklist. Save as `_docs/02_document/FINAL_report.md`.
 
 ---
 

.cursor/skills/problem/SKILL.md

@@ -181,6 +181,8 @@ Categorized measurable criteria with markdown headers and bullet points:
 
 Every criterion must have a measurable value. Vague criteria like "should be fast" are not acceptable — push for "less than 400ms end-to-end".
 
+**AC must be design-independent**: describe testable outcomes only — no libraries, algorithms, params, or design choices. Implementation follows AC, never reverse. (IEEE 830 / Atlassian / GitScrum)
+
 ### input_data/
 
 At least one file. Options:

.cursor/skills/refactor/SKILL.md

@@ -25,6 +25,7 @@ Phase details live in `phases/` — read the relevant file before executing each
 - **Delegate execution**: all code changes go through the implement skill via task files
 - **Ask, don't assume**: when scope or priorities are unclear, STOP and ask the user
 - **Exact-fit recommendations**: do not recommend a replacement pattern, library, service, architecture, algorithm, or "modern approach" merely because it improves structure or solves a similar class of problem. It must fit confirmed product constraints, acceptance criteria, operating context, integration boundaries, and current code realities. Otherwise reject it, mark it experimental, or ask the user before adding it to the roadmap.
+- **Per-mode API capability verification on replacements**: when a refactor proposes replacing or adding a library/SDK/framework/service that exposes multiple modes or configurations, pin the exact mode the refactored code will use (inputs, outputs, runtime) and verify *that mode* via mandatory `context7` lookup plus a saved Minimum Viable Example before promoting the recommendation to `Selected`. Capability claims at the category level ("supports A, B, C modes") must be cross-checked against the literal mode enumeration — `A, B → A+B` style conflations are the recurring silent-failure path.
 
 ## Context Resolution
 
@@ -58,7 +59,7 @@ Create REFACTOR_DIR and RUN_DIR if missing. If a RUN_DIR with the same name alre
 
 Both modes produce `RUN_DIR/list-of-changes.md` (template: `templates/list-of-changes.md`). Both modes then convert that file into task files in TASKS_DIR during Phase 2.
 
-**Guided mode cleanup**: after `RUN_DIR/list-of-changes.md` is created from the input file, delete the original input file to avoid duplication.
+**Guided mode cleanup**: after `RUN_DIR/list-of-changes.md` is created from the input file, delete the original input file only if it lives outside `RUN_DIR`. If the provided file is already the canonical `RUN_DIR/list-of-changes.md`, keep it as the audit record.
 
 ## Workflow
 
@@ -80,10 +81,10 @@ Both modes produce `RUN_DIR/list-of-changes.md` (template: `templates/list-of-ch
 - "refactor [specific target]" → skip phase 1 if docs exist
 - Default → all phases
 
-**Testability-run specifics** (guided mode invoked by autodev existing-code flow Step 4):
+**Testability-run specifics** (guided mode invoked by autodev existing-code Step 4 or greenfield Step 8):
 - Run name is `01-testability-refactoring`.
 - Phase 3 (Safety Net) is skipped by design — no tests exist yet. Compensating control: the `list-of-changes.md` gate in Phase 1 must be reviewed and approved by the user before Phase 4 runs.
-- Scope is MINIMAL and surgical; reject change entries that drift into full refactor territory (see existing-code flow Step 4 for allowed/disallowed lists). Flagged entries go to `RUN_DIR/deferred_to_refactor.md` for Step 8 (optional full refactor) consideration.
+- Scope is MINIMAL and surgical; reject change entries that drift into full refactor territory (see the invoking flow's testability step for allowed/disallowed lists). Flagged entries go to `RUN_DIR/deferred_to_refactor.md` for the next optional full-refactor step or backlog consideration.
 - After Phase 4 (Execution) completes, write `RUN_DIR/testability_changes_summary.md` as Phase 4.5. Format: one bullet per applied change.
   ```markdown
   # Testability Changes Summary ({{run_name}})

.cursor/skills/refactor/phases/02-analysis.md

@@ -10,6 +10,17 @@
 2. Extract the **Project Constraint Matrix** from `problem.md`, `restrictions.md`, `acceptance_criteria.md`, current architecture/docs, and actual code constraints. Include required inputs/outputs, operating context, lifecycle assumptions, integration boundaries, non-functional targets, and hard disqualifiers.
 3. Research modern approaches for similar systems
 4. For each alternative pattern/library/service/architecture/algorithm, research intrinsic implementation constraints: required inputs/outputs, runtime assumptions, supported deployment modes, resource needs, operational limits, licensing/security constraints, and known failure reports.
+
+   **API Capability Verification — Per-Mode (MANDATORY, BLOCKING for proposed replacements)**
+
+   When a refactor recommendation replaces (or adds) a library/SDK/framework/service, the same per-mode verification used by `/research` Step 2 applies — selecting a replacement on category fit alone is the same silent-failure path. For every replacement candidate that has multiple modes or configurations:
+
+   1. **Pin the exact mode/configuration** the refactored code will use, in one explicit sentence. Inputs (data shapes, sensor counts, payloads, rates), outputs (per `acceptance_criteria.md` and contract files), runtime (matching the project's deployment).
+   2. **Run `context7` (or equivalent docs lookup)** for the candidate. **Mandatory for every replacement library/SDK/framework candidate**, not optional. Minimum three queries per candidate: mode enumeration, project's exact mode (with input/output shapes), disqualifier probe ("does this mode produce the required output? are there published limitations on this runtime?"). Append URLs to `RUN_DIR/analysis/research_findings.md` references section.
+   3. **Save a Minimum Viable Example (MVE)** for the pinned mode under `RUN_DIR/analysis/mve_evidence.md` with: source, inputs in example, outputs in example, project inputs, project outputs required, match assessment ✅/⚠️/❌. If no official example covers the project's exact configuration, the recommendation cannot be `Selected` based on category fit alone — it must be `Experimental only` (with required-evidence note) or `Rejected`.
+   4. **Treat "the same library in a different mode" as a different recommendation.** If the project's pinned mode is `<X>` but the only documented evidence covers `<Y>`, do not silently soften the description. Open a separate recommendation row, with its own MVE, fit assessment, and disqualifiers.
+   5. **Common silent-failure pattern**: a fact summary paraphrases docs as "supports A, B, C, D modes" when the docs actually mean "supports A; B; C and D as separate orthogonal modes" — no `A+B` combination exists. Cross-check paraphrased capability claims against the literal mode enumeration.
+
 5. Identify what could be done differently
 6. Suggest improvements only when they fit the Project Constraint Matrix. A cleaner or more modern approach that violates product constraints must be marked `Rejected` or `Experimental only`, not added as a roadmap recommendation.
 
@@ -17,7 +28,8 @@ Write `RUN_DIR/analysis/research_findings.md`:
 - Current state analysis: patterns used, strengths, weaknesses
 - Alternative approaches per component: current vs alternative, pros/cons, migration effort
 - Prioritized recommendations: quick wins + strategic improvements
-- Constraint-fit table: recommendation, constraints checked, evidence, mismatches/disqualifiers, status (`Selected` / `Rejected` / `Experimental only` / `Needs user decision`)
+- Constraint-fit table: recommendation, **pinned mode/config**, constraints checked, **API capability evidence (MVE link)**, evidence, mismatches/disqualifiers, status (`Selected` / `Rejected` / `Experimental only` / `Needs user decision`)
+- For every recommendation that replaces or adds a library/SDK/framework, append a **Restrictions × Candidate-Mode sub-matrix** that walks every numbered line of `restrictions.md` and `acceptance_criteria.md` against the candidate's pinned mode, marking each cell ✅ Pass / ❌ Fail / ❓ Verify / N/A with cited evidence. A recommendation cannot be `Selected` while any cell is ❌ or ❓.
 
 ## 2b. Solution Assessment & Hardening Tracks
 
@@ -27,6 +39,44 @@ Write `RUN_DIR/analysis/research_findings.md`:
 4. Prioritize changes by impact and effort
 5. Reject or escalate any proposed refactor that improves code structure while weakening required behavior, integration contracts, runtime constraints, safety/security posture, or acceptance criteria
 
+### 2b.1. ADR Superseding Gate (BLOCKING)
+
+A refactor that improves code structure while overturning a documented architecture decision is the silent-drift class the project repeatedly burns on (see `meta-rule.mdc` § GPS-passthrough postmortem and the auto-lessons it produced). This gate makes drift visible and forces a deliberate ADR update.
+
+1. **List candidate ADRs**: read every `Status: Accepted` file in `_docs/02_document/adr/`. If the directory does not exist or contains only the index, log `No ADRs in scope` to `RUN_DIR/analysis/adr_impact.md` and skip the rest of this gate.
+2. **Diff each candidate against the proposed refactor roadmap**: for each ADR, ask the same two questions as code-review Phase 7:
+   - **Violation**: does any roadmap item do the *opposite* of the ADR's `Decision`?
+   - **Drift**: does any roadmap item materially affect the ADR's `Consequences` (positive or negative) without contradicting the Decision outright?
+3. **Classify each impacted ADR** in `RUN_DIR/analysis/adr_impact.md`:
+
+   | ADR | Roadmap item | Impact | Required action |
+   |-----|--------------|--------|-----------------|
+   | NNN | `roadmap-item-NN` | Violation / Drift / Aligned | (filled by Choose A/B/C below) |
+
+4. **For every Violation row, present a BLOCKING Choose**:
+
+   ```
+   ══════════════════════════════════════
+    DECISION REQUIRED: Refactor would violate ADR-NNN (<title>)
+   ══════════════════════════════════════
+    A) Update the ADR via supersede: the refactor produces a NEW ADR
+       (`Supersedes: NNN`) capturing the new Decision, and ADR-NNN's
+       `Superseded by` field is updated. The supersede ADR is itself a
+       deliverable of this refactor run (added to RUN_DIR/analysis/adr_impact.md
+       and to TASKS_DIR as a task) and must be `Accepted` before Phase 4.
+    B) Reduce the refactor scope to NOT violate ADR-NNN
+    C) Re-evaluate ADR-NNN: keep the refactor but only after ADR-NNN is
+       formally re-opened in a new /plan Step 4.5 round
+   ══════════════════════════════════════
+    Recommendation: A — supersede is the only path that keeps the audit
+    trail intact while letting the refactor land
+   ══════════════════════════════════════
+   ```
+
+5. **For every Drift row**: do not block, but the roadmap item must include a `## ADR Impact` section in its task spec citing the affected ADR(s). The implementer surfaces this at code-review Phase 7, which would otherwise classify the change as ADR-Drift (High) without context.
+6. **For every Aligned row**: cite the ADR in the roadmap item's task spec under `## ADR Compliance`. No further action.
+7. **Self-supersede deliverable**: any Choose A path adds a `[##]_supersede_adr_NNN.md` task file to the refactor run's TASKS_DIR with the new ADR text drafted (using `.cursor/skills/plan/templates/adr.md`). The task's only Acceptance Criterion is "ADR file exists at `_docs/02_document/adr/<next>_<slug>.md` with `Status: Accepted`, ADR-NNN's `Superseded by` field updated, and `_docs/02_document/adr/README.md` index reflects both."
+
 Present optional hardening tracks for user to include in the roadmap:
 
 ```
@@ -55,6 +105,8 @@ Write `RUN_DIR/analysis/refactoring_roadmap.md`:
 
 **BLOCKING applicability gate**: Before 2c and 2d, every recommendation in the roadmap must be `Selected`. Items marked `Rejected` are excluded. Items marked `Experimental only` or `Needs user decision` require a user decision before task creation.
 
+**BLOCKING ADR-supersede gate**: Before 2c and 2d, every Violation row in `RUN_DIR/analysis/adr_impact.md` (from 2b.1) must be resolved via Choose A, B, or C. A Violation row with no chosen path blocks task creation.
+
 ## 2c. Create Epic
 
 Create a work item tracker epic for this refactoring run:
@@ -62,7 +114,7 @@ Create a work item tracker epic for this refactoring run:
 1. Epic name: the RUN_DIR name (e.g., `01-testability-refactoring`)
 2. Create the epic via configured tracker MCP
 3. Record the Epic ID — all tasks in 2d will be linked under this epic
-4. If tracker unavailable, use `PENDING` placeholder and note for later
+4. If tracker is unavailable, follow `.cursor/rules/tracker.mdc`; only use `PENDING` placeholders if the user explicitly chooses `tracker: local`
 
 ## 2d. Task Decomposition
 
@@ -88,6 +140,9 @@ Convert the finalized `RUN_DIR/list-of-changes.md` into implementable task files
 - [ ] Recommendations are grounded in actual code, not abstract
 - [ ] Every recommendation has been checked against the Project Constraint Matrix
 - [ ] No recommendation violates product restrictions, acceptance criteria, documented architecture decisions, or actual code integration boundaries
+- [ ] Every replacement library/SDK/framework recommendation has a pinned mode/config, a saved MVE in `mve_evidence.md`, and a Restrictions × Candidate-Mode sub-matrix with no ❌ or ❓ cells
+- [ ] `context7` (or equivalent) was consulted for every replacement library/SDK/framework recommendation
+- [ ] Paraphrased capability claims have been cross-checked against the literal mode-enumeration evidence (no `A, B → A+B` style conflation)
 - [ ] Rejected and experimental approaches are documented but not converted into implementation tasks without user approval
 - [ ] Roadmap phases are prioritized by impact
 - [ ] Epic created and all tasks linked to it
@@ -96,6 +151,10 @@ Convert the finalized `RUN_DIR/list-of-changes.md` into implementable task files
 - [ ] Task dependencies are consistent (no circular dependencies)
 - [ ] `_dependencies_table.md` includes all refactoring tasks
 - [ ] Every task has a work item ticket (or PENDING placeholder)
+- [ ] If `_docs/02_document/adr/` exists with Accepted ADRs, `RUN_DIR/analysis/adr_impact.md` has been written and every Violation row is resolved (A/B/C) — no implicit overrides
+- [ ] For every Violation resolved via Choose A, a `[##]_supersede_adr_NNN.md` task exists in TASKS_DIR with the drafted supersede ADR
+- [ ] For every Drift row, the corresponding roadmap-item task spec has a `## ADR Impact` section
+- [ ] For every Aligned row, the corresponding roadmap-item task spec has a `## ADR Compliance` section
 
 **Save action**: Write analysis artifacts to RUN_DIR, task files to TASKS_DIR
 

.cursor/skills/refactor/phases/03-safety-net.md

@@ -15,9 +15,9 @@ Before designing or implementing any new tests, check what already exists:
 1. Scan the project for existing test files (unit tests, integration tests, blackbox tests)
 2. Run the existing test suite — record pass/fail counts
 3. Measure current coverage against the areas being refactored (from `RUN_DIR/list-of-changes.md` file paths)
-4. Assess coverage against thresholds:
+4. Assess coverage against thresholds (canonical: see `.cursor/rules/cursor-meta.mdc` Quality Thresholds — never hardcode a different number):
    - Minimum overall coverage: 75%
-   - Critical path coverage: 90%
+   - Critical path coverage: **90% floor / 100% aim** — 90% is the enforcement floor (blocks Phase 4 if not met); 100% is the aspirational target. Refactors are NOT permitted to drop below 90% on the critical paths covered by the in-scope changes.
    - All public APIs must have blackbox tests
    - All error handling paths must be tested
 
@@ -47,7 +47,7 @@ For each uncovered critical area, write test specs to `RUN_DIR/test_specs/[##]_[
 4. Document any discovered issues
 
 **Self-verification**:
-- [ ] Coverage requirements met (75% overall, 90% critical paths) across existing + new tests
+- [ ] Coverage requirements met (75% overall, 90% critical-path floor — 100% aim — per canonical `cursor-meta.mdc` Quality Thresholds) across existing + new tests
 - [ ] All tests pass on current codebase
 - [ ] All public APIs in refactoring scope have blackbox tests
 - [ ] Test data fixtures are configured

.cursor/skills/refactor/phases/04-execution.md

@@ -10,7 +10,7 @@
    - All `[TRACKER-ID]_refactor_*.md` files are present
    - Each task file has valid header fields (Task, Name, Description, Complexity, Dependencies)
 2. Verify `TASKS_DIR/_dependencies_table.md` includes the refactoring tasks
-3. Verify all tests pass (safety net from Phase 3 is green)
+3. Verify all tests pass (safety net from Phase 3 is green), unless this is a testability run where Phase 3 was intentionally skipped
 4. If any check fails, go back to the relevant phase to fix
 
 ## 4b. Delegate to Implement Skill
@@ -23,7 +23,7 @@ The implement skill will:
 3. Compute execution batches for the refactoring tasks
 4. Implement tasks sequentially in topological order (no subagents, no parallelism)
 5. Run code review after each batch
-6. Commit and push per batch
+6. Commit per batch and push only when the user approved pushing
 7. Update work item ticket status
 
 Do NOT modify, skip, or abbreviate any part of the implement skill's workflow. The refactor skill is delegating execution, not optimizing it.
@@ -47,7 +47,7 @@ After the implement skill completes:
 For each successfully completed refactoring task:
 
 1. Transition the work item ticket status to **Done** via the configured tracker MCP
-2. If tracker unavailable, note the pending status transitions in `RUN_DIR/execution_log.md`
+2. If tracker is unavailable, follow `.cursor/rules/tracker.mdc`; if the user explicitly chose `tracker: local`, note the pending status transitions in `RUN_DIR/execution_log.md`
 
 For any failed or blocked tasks, leave their status as-is (the implement skill already set them to In Testing or blocked).
 

.cursor/skills/refactor/phases/05-test-sync.md

@@ -45,7 +45,7 @@ Write `RUN_DIR/test_sync/new_tests.md`:
 - [ ] All obsolete tests removed or merged
 - [ ] All pre-existing tests pass after updates
 - [ ] New code from Phase 4 has test coverage
-- [ ] Overall coverage meets or exceeds Phase 3 baseline (75% overall, 90% critical paths)
+- [ ] Overall coverage meets or exceeds Phase 3 baseline (75% overall, 90% critical-path floor / 100% aim — per `.cursor/rules/cursor-meta.mdc` Quality Thresholds)
 - [ ] No tests reference removed or renamed code
 
 **Save action**: Write test_sync artifacts; implemented tests go into the project's test folder

.cursor/skills/release/SKILL.md

@@ -0,0 +1,290 @@
+---
+name: release
+description: |
+  Executes the deployment plan produced by /deploy against a target environment.
+  Closes the loop between "we have a plan" and "the new version is running in production with a verdict on disk."
+  6-phase workflow: pre-release gate, strategy select, execute, smoke test, watch window, commit-or-rollback.
+  Outputs _docs/04_release/release_<version>.md with a definitive Released / Rolled-Back / Aborted verdict.
+  Trigger phrases:
+  - "release", "ship", "go live", "release this version"
+  - "deploy to prod", "promote to staging", "roll out"
+  - "rollback", "abort the release"
+category: ship
+tags: [release, deployment, rollback, smoke-test, observability, production]
+disable-model-invocation: true
+---
+
+# Release Execution
+
+The `/deploy` skill produces a plan and scripts. The `/release` skill **runs** them, verifies the live system, watches it for a defined window, and produces a definitive verdict on disk.
+
+## Core Principles
+
+- **Real execution, not simulation**: every phase must actually run against the target environment. If a phase cannot be executed (missing scripts, no SSH access, disabled secrets, registry auth failure), STOP — do not pretend a step succeeded. See `meta-rule.mdc` § "Real Results, Not Simulated Ones".
+- **Verifiable rollback path**: the release does not start until rollback is proven viable for this version. "We can roll back" without evidence is not a rollback path.
+- **Quiet failure is a release failure**: a deploy script that exits 0 but emits no observable signal in the watch window is treated as a regression, not a success.
+- **One release per invocation**: a single `/release` execution targets exactly one version against exactly one environment. Multi-stage promotion (staging → prod) is two invocations, not one.
+- **Never skip the watch window**: even successful deploys can degrade after 5–60 minutes (cache warm-up, scheduled jobs, downstream backpressure). The watch window is mandatory.
+- **Autonomous rollback on hard regressions**: critical health-check failure, error-rate spike above threshold, or smoke-test failure → automatic rollback. Soft regressions (latency drift, capacity warnings) escalate to the user.
+
+## Context Resolution
+
+Fixed paths:
+
+- DEPLOY_DIR: `_docs/04_deploy/`
+- RELEASE_DIR: `_docs/04_release/`
+- SCRIPTS_DIR: `scripts/`
+- DEPLOY_SCRIPT: `scripts/deploy.sh`
+- HEALTH_SCRIPT: `scripts/health-check.sh`
+- ENV_TEMPLATE: `.env.example`
+- OBSERVABILITY_DOC: `_docs/04_deploy/observability.md`
+- ENVIRONMENT_DOC: `_docs/04_deploy/environment_strategy.md`
+- PROCEDURES_DOC: `_docs/04_deploy/deployment_procedures.md`
+- ARCHITECTURE: `_docs/02_document/architecture.md`
+- RESTRICTIONS: `_docs/00_problem/restrictions.md`
+
+Announce the resolved paths and the **target environment + version + strategy** to the user before any phase that touches the live system.
+
+## Inputs (BLOCKING prerequisites)
+
+| Input | Required | Source |
+|-------|----------|--------|
+| Target environment | Yes — ASK user | `environment_strategy.md` enumerates valid options |
+| Target version / image tag | Yes — ASK user | Must exist in the registry; verified in Phase 1 |
+| Rollback target version | Yes — ASK user | Defaults to currently-deployed version if discoverable |
+| `scripts/deploy.sh` | Yes | Produced by `/deploy` Step 7. STOP if missing → run `/deploy` first |
+| `scripts/health-check.sh` | Yes | Same |
+| `_docs/04_deploy/deployment_procedures.md` | Yes | Defines per-environment runbook, manual approval rules, change-window restrictions |
+| `_docs/04_deploy/observability.md` | Yes | Defines watch metrics, thresholds, and dashboards |
+| `_docs/04_deploy/environment_strategy.md` | Yes | Defines target hostnames, registries, secrets, deploy strategy per env |
+
+## Outputs
+
+```
+RELEASE_DIR/
+├── release_<version>_<env>_<YYYY-MM-DD-HHmm>.md       (mandatory; one per invocation)
+├── rollback_<version>_<env>_<YYYY-MM-DD-HHmm>.md      (only when rollback fires; pairs with the release file)
+└── manual_approvals/
+    └── approval_<version>_<env>.md                     (when restrictions require manual approval, written before Phase 3)
+```
+
+The release report (`templates/release-report.md`) is appended to as each phase completes — it is durable across phase failures and reflects partial progress so the next operator can resume or audit.
+
+## Phases
+
+```
+┌────────────────────────────────────────────────────────────────┐
+│             Release Execution (6-Phase Method)                  │
+├────────────────────────────────────────────────────────────────┤
+│ PREREQ: deploy artifacts on disk; tests green at HEAD          │
+│                                                                │
+│ 1. Pre-Release Gate     → AC + change summary + readiness      │
+│    [BLOCKING: user confirms or aborts]                         │
+│ 2. Strategy Select      → all-at-once / blue-green / canary    │
+│    [BLOCKING: user picks strategy]                             │
+│ 3. Execute              → run deploy.sh, capture exit + logs   │
+│    [AUTO-ROLLBACK on non-zero exit]                            │
+│ 4. Smoke Test           → /test-run prod-smoke in target env   │
+│    [AUTO-ROLLBACK on failure]                                  │
+│ 5. Watch Window         → poll observability for N minutes     │
+│    [AUTO-ROLLBACK on hard threshold breach]                    │
+│ 6. Commit or Rollback   → finalize verdict, update tracker     │
+│    [BLOCKING: user confirms only if soft regression escalated] │
+├────────────────────────────────────────────────────────────────┤
+│ Verdicts: Released · Rolled-Back · Aborted                     │
+└────────────────────────────────────────────────────────────────┘
+```
+
+### Phase 1: Pre-Release Gate
+
+**Goal**: Refuse to start if the system is not ready for a real release.
+
+1. **Acceptance criteria check**: read `_docs/00_problem/acceptance_criteria.md`. If any AC is marked unmet OR if any AC has no associated test marked `Passed` in the latest `test-run` report, STOP and surface the unmet items. Do not let the user override with "ship anyway" without a recorded reason in the release report.
+2. **Test status check**: read the most recent `_docs/06_metrics/perf_*.md` (if perf is required by restrictions) and the latest functional test report. Any failing or skipped test that maps to a critical-path AC blocks the release.
+3. **Change summary**: read the git log between the version-tag-of-last-release and HEAD (or, if no prior release exists, from the project root commit). Render a short list grouped by component: features, fixes, breaking changes, security fixes. Cross-reference against the latest implementation reports under `_docs/03_implementation/`.
+4. **Rollback readiness**:
+   - Confirm the previous version's image is still pullable from the registry (do not deploy without this).
+   - Confirm `scripts/deploy.sh --rollback` works as documented (read the script; if `--rollback` flag is missing, STOP — that is a deploy-skill bug).
+   - Confirm a rollback target exists (e.g., previously-deployed image tag) and is recorded in the release report under `Rollback Plan`.
+5. **Restrictions**: read `_docs/00_problem/restrictions.md` for change-window rules, manual-approval rules, blackout windows, regulatory requirements (e.g., 4-eyes review, ITAR controls). If any apply, gate accordingly — write a `manual_approvals/approval_<version>_<env>.md` file once received.
+6. **Tracker check**: list tracker tickets in the release scope (per `tracker.mdc` rules). Any ticket still in `In Progress` or `Code Review` that maps to a change in the release scope blocks Phase 1. Move-and-deploy is not allowed.
+
+**BLOCKING gate**: present the assembled summary to the user using Choose A/B/C:
+
+```
+══════════════════════════════════════
+ PRE-RELEASE GATE
+══════════════════════════════════════
+ Target env:        {env}
+ Target version:    {version} ({git-sha})
+ Rollback target:   {previous-version}
+ Changes:           N tickets, M components
+   - {summary list}
+ Open risks:        {summary or "none"}
+ Blocking issues:   {summary or "none"}
+══════════════════════════════════════
+ A) Proceed to Strategy Select
+ B) Abort — fix blocking issue and re-invoke
+ C) Edit release scope — exclude a ticket and reassemble
+══════════════════════════════════════
+```
+
+If A → write Phase 1 section to release report, proceed. If B → write `Aborted` verdict to release report with reason, exit. If C → loop back into Phase 1 with edited scope.
+
+### Phase 2: Strategy Select
+
+**Goal**: Pick the deployment strategy that fits the change risk and environment capability.
+
+Read `environment_strategy.md` and `deployment_procedures.md` to learn which strategies the target env supports. Strategies and when each is appropriate:
+
+| Strategy | When to pick | Risk if wrong |
+|----------|--------------|---------------|
+| **all-at-once** | Internal tools, low traffic, well-rehearsed change, env supports nothing else | All users hit the new version simultaneously — bug blast radius is 100% |
+| **blue-green** | Stateless services with a load balancer, env has dual-stack capability | Cutover is binary — observability must be ready to detect issues fast |
+| **canary** | Customer-facing, traffic-tier load balancer in place, gradual rollout possible | Canary metric thresholds must be well-tuned or canary fails for harmless reasons |
+| **manual** | Non-automatable env (one-off VMs, regulated infrastructure, non-Docker host) | The whole release becomes a runbook and the watch window phases are operator-driven; the release skill records but does not execute |
+
+Recommend a default based on:
+- Risk level inferred from change summary (any breaking change → bias toward canary or blue-green)
+- Restrictions (e.g., regulatory rules forcing manual approval at each step)
+- Environment capability (some envs may only support all-at-once)
+
+**BLOCKING gate**: Choose A/B/C/D between strategies. Record the choice in the release report.
+
+### Phase 3: Execute
+
+**Goal**: Actually run the deploy. Capture exit code and full stdout/stderr.
+
+1. Validate environment file (`.env`) exists, all required vars from `.env.example` are set, no placeholder secrets remain.
+2. Source the env file and run `scripts/deploy.sh` against the target host. The script produced by `/deploy` Step 7 is the point of execution; do NOT bypass it. If a strategy-specific flag is needed (e.g., `--canary 5%`), pass it through.
+3. Stream stdout/stderr to the release report, with timestamps, in a fenced code block under `## Phase 3: Execute`.
+4. Capture exit code.
+5. **AUTO-ROLLBACK trigger**: non-zero exit code → immediately invoke Phase 6 with verdict `Rolled-Back: deploy script failure`. Do NOT continue to Phase 4.
+
+If `deploy.sh` emits no output for more than the configured idle threshold (default 5 minutes; check `deployment_procedures.md` for an explicit value), treat it as hung — capture a snapshot of what's running on the target, kill the script, and AUTO-ROLLBACK with reason `Deploy hung — manual investigation required`.
+
+**Manual strategy**: if Phase 2 picked `manual`, write a checklist of operator steps from `deployment_procedures.md` to the release report and pause until the user types `done` or `failed`. Phase 3 then records the user's report verbatim.
+
+### Phase 4: Smoke Test
+
+**Goal**: Verify the new version is *actually serving traffic correctly* in the target environment.
+
+1. Resolve the smoke-test command from `_docs/02_document/tests/blackbox-tests.md` § Production Smoke Tests, OR delegate to `/test-run` in `--prod-smoke` mode against the target environment.
+2. The smoke-test set must (a) hit each public endpoint of each component, (b) include at least one read AND one write per public endpoint where applicable, and (c) complete in under 5 minutes total.
+3. Capture pass/fail per case to the release report.
+4. **AUTO-ROLLBACK trigger**: any smoke-test failure → invoke Phase 6 with verdict `Rolled-Back: smoke test failure: <test-name>`.
+
+If smoke tests are **missing** for the target environment (no production-mode test set), STOP — write a leftover entry to `_docs/_process_leftovers/` per `tracker.mdc`, do not proceed to watch window without smoke coverage. Write `Aborted: smoke tests missing for prod-mode target` and ASK the user.
+
+### Phase 5: Watch Window
+
+**Goal**: Observe the live system for a defined window to catch latent regressions.
+
+1. Read `observability.md` for the project's metrics, dashboards, and threshold definitions. Required watch metrics for any production target (per cursor-meta convention) include error rate, request rate, p99 latency, and saturation (CPU/memory/queue-depth).
+2. Compute the watch-window duration from `deployment_procedures.md`. If unspecified, default to **15 minutes** for staging and **60 minutes** for production.
+3. Poll the observability backend at 1-minute intervals (or the configured cadence). For each interval, record metric snapshots to the release report.
+4. Threshold rules:
+   - **Hard breach** (auto-rollback): error-rate ≥ 2× baseline, p99 latency ≥ 3× baseline, any health-check failure persisting for 2 consecutive intervals.
+   - **Soft breach** (escalate): metric drift between 1.5× and 2× baseline, single-interval health blip, queue-depth steady but elevated.
+   - **No data** (escalate): if metrics are not flowing within the first 3 minutes, treat the absence as a hard breach — observability is itself broken.
+5. **AUTO-ROLLBACK trigger**: hard breach at any interval. Move to Phase 6 with verdict `Rolled-Back: <metric> breached <multiplier>× baseline at T+<minutes>`.
+6. **ESCALATE trigger**: soft breach. Pause polling, surface the metric, and ask the user A/B/C:
+   - A) Continue watch — accept current drift, keep polling
+   - B) Roll back now — treat soft drift as hard
+   - C) Extend watch window by N minutes
+7. End of watch window with no breach → proceed to Phase 6.
+
+The watch window cannot be skipped. If the user explicitly demands skipping (e.g., emergency rollforward), record the override reason in the release report and continue, but mark the verdict as `Released-with-override` — this triggers an automatic incident retrospective per `retrospective/SKILL.md`.
+
+### Phase 6: Commit or Rollback
+
+**Goal**: Finalize the release with a definitive verdict on disk.
+
+**Path A — Commit (clean release)**:
+1. Update tracker tickets: every ticket in scope moves to `Released` (or `Done`, per project convention defined in `tracker.mdc` / `_docs/_repo-config.yaml`).
+2. Tag the git HEAD with `release/<version>` (or the project's tag convention from `deployment_procedures.md`).
+3. Write the final `Released` verdict to the release report with a summary table.
+4. Trigger `/retrospective --cycle-end` with this release as the cycle terminus.
+5. Auto-chain to autodev's next step (Retrospective in greenfield, or feature-cycle loop start in existing-code).
+
+**Path B — Rollback (auto-fired or user-elected)**:
+1. Run `scripts/deploy.sh --rollback` with the rollback target captured in Phase 1.
+2. Stream output to a new file `RELEASE_DIR/rollback_<version>_<env>_<YYYY-MM-DD-HHmm>.md` AND append a summary to the original release report under `## Rollback`.
+3. Re-run Phase 4 (smoke test) and a 5-minute mini watch window against the rolled-back version. If THAT also fails, escalate immediately — the system is in an unknown state and needs human takeover.
+4. Update tracker tickets back to `Ready for Release` (or the project's pre-release status).
+5. Write the final `Rolled-Back` verdict with full reason chain.
+6. Auto-trigger `/retrospective --incident` with this release as the incident anchor (per `retrospective/SKILL.md` incident mode).
+7. Do NOT auto-chain to anything else — the user owns the next step.
+
+**Path C — Aborted**:
+Reached only via Phase 1 Choose B, Phase 4 smoke-tests-missing escalation, or any phase that detects a precondition violation. Write `Aborted: <reason>` to the release report. Do not auto-chain.
+
+## Self-verification
+
+- [ ] Release report exists at `RELEASE_DIR/release_<version>_<env>_<timestamp>.md` with verdict (Released / Rolled-Back / Aborted)
+- [ ] Every phase that ran has a section in the release report with timestamps and tool output
+- [ ] On Released: tracker tickets moved to release status; git tag pushed (if convention)
+- [ ] On Rolled-Back: rollback report exists at `RELEASE_DIR/rollback_<version>_<env>_<timestamp>.md`; tracker tickets moved back to pre-release status; incident retrospective scheduled
+- [ ] On Aborted: reason recorded; no live-system changes attempted; no tracker movement
+- [ ] No phase was skipped without an explicit reason recorded in the release report
+
+## Escalation Rules
+
+| Situation | Action |
+|-----------|--------|
+| `scripts/deploy.sh` missing or `--rollback` unsupported | STOP — return to `/deploy` Step 7, do not patch the script in `/release` |
+| Registry auth failure during pre-release | STOP — fix credentials at infra layer (per `coderule.mdc`); do not embed creds in the script |
+| Smoke tests missing for prod target | STOP — write a leftover; do not improvise smoke tests in `/release` |
+| Observability backend unreachable | STOP — observability blindness is itself a release blocker |
+| User asks to skip the watch window | Record override, mark verdict `Released-with-override`, fire incident retro |
+| Rollback also fails its smoke test | ESCALATE to user — system is in unknown state; do not loop deploys |
+| Tracker MCP returns Unauthorized during ticket movement | Per `tracker.mdc`, write a leftover entry; do NOT silently continue without confirming the move |
+| Multiple environments named in user request | STOP — one release per invocation; ask user to pick one |
+| Production smoke test would touch real customer data | STOP — that is a `coderule.mdc` violation; ask user to define a smoke endpoint or test account |
+
+## Common Mistakes
+
+- **Skipping the watch window when "everything looks fine after deploy"** — a deploy that exited 0 is not a release that's stable. Watch is mandatory.
+- **Faking smoke tests** to pass the gate when the prod test set is incomplete. STOP and surface the gap; do not embed prod URLs into ad-hoc curl commands.
+- **Rolling forward through a failure** ("the next deploy will fix it"). Roll back first, fix the cause, then deploy a real fix.
+- **Treating the release report as optional** when only an internal tool changed. Every release writes a report — the audit trail is the value, not the prose volume.
+- **Approving manual gates yourself** without the user's input when restrictions require human approval. The release skill records, the human approves.
+- **Reusing `release_<version>` filenames** across attempted releases. Always include the timestamp in the filename so re-attempts are visible side-by-side.
+- **Letting tracker drift silently** between release attempts. If Phase 6 cannot move tickets, the release is not complete — write a leftover and stop.
+
+## Project Mode vs Standalone
+
+- **Project mode** (default): autodev invokes `/release` after `/deploy`. State writes occur under `_docs/_autodev_state.md`. Full integration with retrospective and feature-cycle loop.
+- **Standalone mode**: `/release` invoked directly with `@<artifact>` (rare; usually only for re-running a rollback against a specific version). All outputs still go to `RELEASE_DIR/`.
+
+## Methodology Quick Reference
+
+```
+┌────────────────────────────────────────────────────────────────┐
+│                Release (6 phases, 3 verdicts)                  │
+├────────────────────────────────────────────────────────────────┤
+│ Phase 1  Pre-Release Gate                                      │
+│           AC + tests + change summary + rollback path          │
+│           [BLOCKING — user A/B/C]                              │
+│ Phase 2  Strategy Select                                       │
+│           all-at-once · blue-green · canary · manual           │
+│           [BLOCKING — user picks]                              │
+│ Phase 3  Execute                                               │
+│           scripts/deploy.sh, capture exit code + logs          │
+│           [AUTO-ROLLBACK on non-zero or hang]                  │
+│ Phase 4  Smoke Test                                            │
+│           /test-run --prod-smoke against target                │
+│           [AUTO-ROLLBACK on any failure]                       │
+│ Phase 5  Watch Window                                          │
+│           Poll observability for N minutes                     │
+│           [AUTO-ROLLBACK on hard breach; escalate on soft]     │
+│ Phase 6  Commit or Rollback                                    │
+│           Released → tracker, tag, retrospective               │
+│           Rolled-Back → tracker reset, incident retrospective  │
+│           Aborted → no live-system change                      │
+├────────────────────────────────────────────────────────────────┤
+│ Principles: real execution · verifiable rollback ·             │
+│             quiet failure = release failure ·                  │
+│             watch window mandatory                             │
+└────────────────────────────────────────────────────────────────┘
+```

.cursor/skills/release/templates/release-report.md

@@ -0,0 +1,114 @@
+# Release Report — {version} → {env}
+
+- **Date**: {YYYY-MM-DD HH:MM} {timezone}
+- **Operator**: {user}
+- **Strategy**: {all-at-once | blue-green | canary | manual}
+- **Verdict**: {Released | Released-with-override | Rolled-Back | Aborted}
+- **Verdict reason**: {one-line summary}
+
+## Pre-Release Gate (Phase 1)
+
+### Acceptance Criteria
+
+| AC ID | Status | Evidence |
+|-------|--------|----------|
+| AC-001 | Met / Unmet | path:section, test report, etc. |
+
+### Test Status
+
+| Suite | Pass | Fail | Skip | Source |
+|-------|------|------|------|--------|
+| Functional | N | N | N | _docs/03_implementation/{batch}.md |
+| Performance | N | N | N | _docs/06_metrics/perf_*.md |
+
+### Change Summary
+
+| Component | Tickets | Type |
+|-----------|---------|------|
+| {component} | TKT-001, TKT-002 | feature / fix / breaking / security |
+
+### Rollback Plan
+
+- Previous version: `{previous-version}` (registry digest: `{sha}`)
+- Rollback script: `scripts/deploy.sh --rollback`
+- Rollback target verified pullable: yes / no
+- Rollback target verified bootable in target env: yes / no
+
+### Restrictions / Approvals
+
+- Change-window restrictions: {none | description}
+- Manual approvals required: {none | reference to approval file}
+
+### Tracker State at Gate
+
+- Tickets in scope: {N}
+- Tickets blocking release: {0 — list any}
+
+## Strategy Select (Phase 2)
+
+- Recommended: {strategy} — reasoning
+- Chosen: {strategy} — reasoning (if differs from recommended)
+
+## Execute (Phase 3)
+
+- Start: {timestamp}
+- End: {timestamp}
+- Exit code: {0 / non-zero}
+
+```
+<scripts/deploy.sh stdout/stderr stream, with timestamps>
+```
+
+## Smoke Test (Phase 4)
+
+- Mode: {/test-run --prod-smoke | manual smoke set}
+- Start: {timestamp}
+- End: {timestamp}
+
+| Test | Result | Notes |
+|------|--------|-------|
+| {name} | Pass / Fail | response time, status, etc. |
+
+## Watch Window (Phase 5)
+
+- Duration: {minutes}
+- Cadence: {minutes per poll}
+- Backend: {observability source — Prometheus, CloudWatch, Datadog, etc.}
+
+| T+min | error_rate | rps | p99_latency | saturation | health | notes |
+|-------|------------|-----|-------------|------------|--------|-------|
+| 0 | … | … | … | … | OK | … |
+| 1 | … | … | … | … | OK | … |
+| … | … | … | … | … | … | … |
+
+### Threshold breaches
+
+- {None | "p99 latency 1.7× baseline at T+8 — soft breach, user accepted continuation"}
+
+## Commit or Rollback (Phase 6)
+
+### If Released
+
+- Tracker tickets moved: {list}
+- Git tag pushed: {tag} → {sha}
+- Retrospective scheduled: yes — {/retrospective --cycle-end output path}
+
+### If Rolled-Back
+
+- Trigger: {auto / user-elected}
+- Reason: {phase + one-line cause}
+- Rollback start: {timestamp}
+- Rollback end: {timestamp}
+- Post-rollback smoke: pass / fail
+- Tracker tickets moved back: {list}
+- Incident retrospective scheduled: yes — {/retrospective --incident output path}
+
+### If Aborted
+
+- Phase that aborted: {1 / 2 / 3 / 4 / 5}
+- Reason: {one-line cause}
+- No live-system changes attempted: yes / no (if live changes, document under Phase 3 above and treat as Rolled-Back instead)
+
+## Lessons (one-liners; full incident retro if Rolled-Back / Released-with-override)
+
+- {Optional: short one-liner observations the operator wants the next /retrospective to consider}

.cursor/skills/research/SKILL.md

@@ -30,7 +30,27 @@ Transform vague topics raised by users into high-quality, deliverable research r
 - **Internet-first investigation** — do not rely on training data for factual claims; search the web extensively for every sub-question, rephrase queries when results are thin, and keep searching until you have converging evidence from multiple independent sources
 - **Multi-perspective analysis** — examine every problem from at least 3 different viewpoints (e.g., end-user, implementer, business decision-maker, contrarian, domain expert, field practitioner); each perspective should generate its own search queries
 - **Question multiplication** — for each sub-question, generate multiple reformulated search queries (synonyms, related terms, negations, "what can go wrong" variants, practitioner-focused variants) to maximize coverage and uncover blind spots
+- **Component option breadth** — for every component area, build a broad option landscape before selecting. Search direct candidates, adjacent-domain alternatives, commercial/open-source variants, classical/simple baselines, current SOTA, and "do not use" failure cases. A component may not be narrowed to one candidate until alternatives have been searched and rejected with evidence.
+- **Component research depth** — for every serious component candidate, go beyond discovery pages. Read official docs, repository/license files, issue discussions, benchmarks, deployment guides, version/platform requirements, security notes, maintenance signals, and real-world failure reports. Extract evidence for inputs/outputs, lifecycle assumptions, runtime/storage/latency fit, integration boundaries, licensing, operational risks, and unsupported scenarios before assigning any selection status.
 - **Exact-fit component selection** — never select a component, tool, library, service, architecture pattern, or algorithm merely because it solves a similar class of problem. It must be proven compatible with the project's explicit operating context, constraints, required inputs/outputs, non-functional requirements, lifecycle assumptions, and acceptance criteria. If fit is unproven or mismatched, mark it `Rejected`, `Experimental only`, or escalate for user decision before it can shape the solution.
+- **Per-mode API capability verification** *(applies only to technical-component selection — see Research Output Class below)* — when a candidate library/SDK/framework/service exposes multiple modes or configurations, *the candidate is not a single thing*. Pin the exact mode the project will use (one explicit sentence: inputs, outputs, runtime), and verify *that mode* against the project's required inputs/outputs via official docs (mandatory `context7` lookup) plus a saved Minimum Viable Example. Capability claims at the category level ("supports X, Y, Z modes") must be cross-checked against the literal mode enumeration before being treated as project-applicable. Two modes of one library are two distinct candidates for the purposes of the Component Applicability Gate. Does not apply to non-technical research (concept comparison, market/policy investigation, knowledge organization, etc.).
+
+## Research Output Class (BLOCKING — set in Step 1)
+
+Before applying any of the technical-component gates (per-mode API capability verification, Component Applicability Gate, Restrictions × Candidate-Mode sub-matrix, MVE evidence, mandatory `context7` lookup), classify the research output into one of two classes. Record the decision in `00_question_decomposition.md` once, near the top, so every downstream step honors it.
+
+| Class | What the output recommends or selects | Examples | Technical-component gates apply? |
+|-------|---------------------------------------|----------|----------------------------------|
+| **Technical-component selection** | One or more libraries, SDKs, frameworks, services, protocols, data formats, infrastructure patterns, algorithms, or APIs that will be implemented or operated against | "Pick a vector database", "Compare auth-token strategies for our API", "Should we use Kafka or RabbitMQ?", architecture / tech-stack / migration drafts (Mode A, Mode B) | **Yes — all gates active** |
+| **Non-technical investigation** | Concept comparisons, knowledge organization, root-cause investigation of an event, market/policy/regulatory/social analysis, literature review, decision support without committing to specific tooling | "Why did adoption stall in Q3?", "Compare phenomenology vs constructivism", "Map regulatory landscape for X", "What do practitioners say about onboarding under remote-first orgs?" | **No — skip API/MVE/sub-matrix gates; the rest of the 8-step engine still applies** |
+
+How to decide:
+1. Inspect the question and the input files (`problem.md`, `restrictions.md`, `acceptance_criteria.md`, or the standalone input file).
+2. If the deliverable will name specific software/services/protocols that someone will then build with or operate, it is **Technical-component selection**.
+3. If the deliverable is a report, comparison, or recommendation that does not commit to specific tooling, it is **Non-technical investigation**.
+4. **Mixed runs are valid.** Some research questions have a non-technical core but include one technical sub-question (or vice versa). In that case classify per component area within the run, not the run as a whole, and note in `00_question_decomposition.md` which component areas trigger the technical-component gates.
+
+When the run is purely **Non-technical investigation**, the rest of the research engine — question decomposition, perspective rotation, exhaustive web search, fact extraction, comparison framework, reasoning chain, validation, deliverable formatting — still applies in full. The sections that get skipped are explicitly the technical gates listed in the table above.
 
 ## Context Resolution
 

.cursor/skills/research/references/quality-checklists.md

@@ -27,14 +27,25 @@
 - [ ] Iterative deepening completed: follow-up questions from initial findings were searched
 - [ ] No sub-question relies solely on training data without web verification
 
+## Component Option Breadth
+
+- [ ] `00_question_decomposition.md` contains a Component Option Search Plan
+- [ ] Every component area was searched across simple baseline, established production, open-source, commercial/vendor, current SOTA, adjacent-domain, no-build/defer, and known-bad options where applicable
+- [ ] Every component area has at least 3 realistic candidates, or a documented explanation of why broad searches found fewer
+- [ ] Each lead candidate has official/source-of-truth evidence plus independent validation when available
+- [ ] Each component area includes at least one baseline/fallback option and at least one rejected or experimental option when possible
+- [ ] Alternative names, synonyms, and neighboring-domain terms were searched before declaring the option landscape complete
+- [ ] Licensing, runtime, platform, maintenance, and unsupported-scenario searches were performed for every lead, fallback, and rejected candidate
+
 ## Mode A Specific
 
 - [ ] Phase 1 completed: AC assessment was presented to and confirmed by user
 - [ ] AC assessment consistent: Solution draft respects the (possibly adjusted) acceptance criteria and restrictions
 - [ ] Competitor analysis included: Existing solutions were researched
 - [ ] All components have comparison tables: Each component lists alternatives with tools, advantages, limitations, security, cost
+- [ ] Component options are broad: component tables include baseline, production, open-source, commercial/vendor, SOTA/research, adjacent-domain, defer/no-build, and disqualified options where applicable
 - [ ] Tools/libraries verified: Suggested tools actually exist and work as described
-- [ ] Component fit matrix completed: `06_component_fit_matrix.md` exists and every selected component/tool/pattern is marked `Selected`
+- [ ] Component fit matrix completed: `06_component_fit_matrix.md` (or `06_component_fit_matrix/` if split) exists and every selected component/tool/pattern is marked `Selected`
 - [ ] No field-adjacent substitution: no selected candidate is chosen only because it solves a similar class of problem while failing the project's explicit constraints
 - [ ] Testing strategy covers AC: Tests map to acceptance criteria
 - [ ] Tech stack documented (if Phase 3 ran): `tech_stack.md` has evaluation tables, risk assessment, and learning requirements
@@ -47,6 +58,7 @@
 - [ ] New draft is self-contained: Written as if from scratch, no "updated" markers
 - [ ] Performance column included: Mode B comparison tables include performance characteristics
 - [ ] Previous draft issues addressed: Every finding in the table is resolved in the new draft
+- [ ] Existing selected components were challenged against a broad alternative landscape before being kept
 - [ ] Existing component fit audited: every old and new component/tool/pattern was checked against `restrictions.md`, `acceptance_criteria.md`, and the Project Constraint Matrix
 - [ ] Rejected/experimental candidates are not lead recommendations unless the user explicitly accepted the risk
 
@@ -68,7 +80,7 @@ When the research topic has Critical or High sensitivity level:
 ## Target Audience Consistency Check (BLOCKING)
 
 - [ ] Research boundary clearly defined: `00_question_decomposition.md` has clear population/geography/timeframe/level boundaries
-- [ ] Every source has target audience annotated in `01_source_registry.md`
+- [ ] Every source has target audience annotated in `01_source_registry.md` (or category files under `01_source_registry/` if split)
 - [ ] Mismatched sources properly handled (excluded, annotated, or marked reference-only)
 - [ ] No audience confusion in fact cards: Every fact has target audience consistent with research boundary
 - [ ] No audience confusion in the report: Policies/research/data cited have consistent target audiences
@@ -84,6 +96,7 @@ When the research topic has Critical or High sensitivity level:
 ## Exact-Fit Validation (BLOCKING)
 
 - [ ] Project Constraint Matrix extracted from problem context before component selection
+- [ ] Component fit matrix includes `Component Area`, `Option Family`, and `Pinned Mode/Config` columns
 - [ ] Every selected component/tool/library/service/pattern/algorithm has evidence for required inputs/outputs and integration boundaries
 - [ ] Every selected candidate has evidence for the operating context and lifecycle assumptions it must support
 - [ ] Every selected candidate has evidence for non-functional targets that are binding for the project
@@ -91,3 +104,21 @@ When the research topic has Critical or High sensitivity level:
 - [ ] Mismatches are recorded as disqualifiers, not softened into generic limitations
 - [ ] Any candidate with unproven fit is marked `Experimental only` or escalated for user decision
 - [ ] Any candidate with documented constraint conflict is marked `Rejected`
+
+## API Capability Verification (BLOCKING)
+
+**Applicability**: this checklist applies only when the run is classified as **Technical-component selection** (see SKILL.md → Research Output Class). For non-technical research (concept comparison, market/policy investigation, root-cause analysis, knowledge organization), skip this checklist entirely and note the skip in `05_validation_log.md`. For mixed runs, apply only to technical component areas.
+
+For every lead candidate that is a library/SDK/framework/service:
+
+- [ ] The exact mode/configuration the project will use is pinned in one explicit sentence (inputs, outputs, runtime); no vague "supports X" language
+- [ ] `context7` (or equivalent docs lookup) was run for the candidate, with at least 3 queries: mode enumeration, project's exact mode, disqualifier probe
+- [ ] All consulted URLs from context7 / official docs are appended to `01_source_registry.md` (or files under `01_source_registry/` if split)
+- [ ] A Minimum Viable Example (MVE) was saved for the pinned mode in `02_fact_cards.md` / `02_fact_cards/` (or `02_mve_evidence.md`) with: source, inputs in example, outputs in example, project inputs, project outputs required, match assessment ✅/⚠️/❌
+- [ ] When the MVE inputs or outputs do not exactly match the project's, the mismatch is cited from the official docs (not inferred), and the candidate is `Experimental only` or `Rejected`
+- [ ] When a library has multiple modes, each project-relevant mode appears as its own candidate row (not a single library row that softens across modes)
+- [ ] Restrictions × Candidate-Modes sub-matrix in `06_component_fit_matrix.md` (or files under `06_component_fit_matrix/` if split) is filled for every lead candidate, with one row per numbered restriction and per numbered acceptance criterion
+- [ ] Sub-matrix uses ✅ / ❌ / ❓ / N/A only — no free-form prose substitutes
+- [ ] No `Selected` candidate has any ❌ or ❓ cell in its sub-matrix
+- [ ] "Validation gate required" footnotes are explicitly classified as either *API capability* (must be resolved here) or *runtime quality* (may be carried forward)
+- [ ] Paraphrased capability claims in fact cards have been cross-checked against the literal mode-enumeration evidence (no `mono, inertial → mono-inertial` style conflation)

.cursor/skills/research/references/source-tiering.md

@@ -89,7 +89,7 @@ Value Translation:
 
 ## Source Registry Entry Template
 
-For each source consulted, immediately append to `01_source_registry.md`:
+For each source consulted, immediately append to `01_source_registry.md` (or the appropriate category file under `01_source_registry/` if the artifact has been split — see splittable-artifacts convention in `steps/00_project-integration.md`):
 ```markdown
 ## Source #[number]
 - **Title**: [source title]

.cursor/skills/research/steps/00_project-integration.md

@@ -63,18 +63,43 @@ RESEARCH_DIR/
     └── source_2.md
 ```
 
+#### Splittable artifacts — Layout convention
+
+The following three artifacts MAY equivalently be a **folder** of the same base name when the single-file form has grown unwieldy (typically ≳ 1000 lines or ≳ 200 KB):
+
+- `01_source_registry.md` ↔ `01_source_registry/`
+- `02_fact_cards.md` ↔ `02_fact_cards/`
+- `06_component_fit_matrix.md` ↔ `06_component_fit_matrix/`
+
+When using the folder form:
+
+- Place a `00_summary.md` index file at the folder root with a short common summary table and the cross-cutting status the single-file form would have carried in its preamble.
+- Split per-entry content into category files (e.g. one file per sub-question or per component): `SQ1_*.md`, `C1_*.md`, etc. Keep entry numbering global across the folder so cross-references like "Source #42" still resolve to exactly one place.
+- Cross-references from outside the folder may point at either `01_source_registry/00_summary.md` (for the index) or directly at the relevant category file.
+
+```
+RESEARCH_DIR/01_source_registry/        # split form (when single-file is too large)
+├── 00_summary.md                       # index + investigation status + compact source table
+├── SQ1_existing_systems.md             # category file
+├── SQ2_canonical_pipeline.md           # category file
+├── C1_vio.md                           # per-component file
+└── ...
+```
+
+Throughout the rest of this skill (other steps, references, templates), the singular `XX.md` form is used as a logical name; treat each occurrence as applying equally to the folder form when the artifact has been split.
+
 ### Save Timing & Content
 
 | Step | Save immediately after completion | Filename |
 |------|-----------------------------------|----------|
 | Mode A Phase 1 | AC & restrictions assessment tables | `00_ac_assessment.md` |
 | Step 0-1 | Question type classification + sub-question list | `00_question_decomposition.md` |
-| Step 2 | Each consulted source link, tier, summary | `01_source_registry.md` |
-| Step 3 | Each fact card (statement + source + confidence) | `02_fact_cards.md` |
+| Step 2 | Each consulted source link, tier, summary | `01_source_registry.md` *(splittable, see convention)* |
+| Step 3 | Each fact card (statement + source + confidence) | `02_fact_cards.md` *(splittable, see convention)* |
 | Step 4 | Selected comparison framework + initial population | `03_comparison_framework.md` |
 | Step 6 | Reasoning process for each dimension | `04_reasoning_chain.md` |
 | Step 7 | Validation scenarios + results + review checklist | `05_validation_log.md` |
-| Step 7.5 | Component exact-fit gate and selection status | `06_component_fit_matrix.md` |
+| Step 7.5 | Component exact-fit gate and selection status | `06_component_fit_matrix.md` *(splittable, see convention)* |
 | Step 8 | Complete solution draft | `OUTPUT_DIR/solution_draft##.md` |
 
 ### Save Principles
@@ -92,12 +117,12 @@ RESEARCH_DIR/
 |------|---------|----------------|
 | `00_ac_assessment.md` | AC & restrictions assessment (Mode A only) | After Phase 1 completion |
 | `00_question_decomposition.md` | Question type, sub-question list | After Step 0-1 completion |
-| `01_source_registry.md` | All source links and summaries | Continuously updated during Step 2 |
-| `02_fact_cards.md` | Extracted facts and sources | Continuously updated during Step 3 |
+| `01_source_registry.md` *(splittable)* | All source links and summaries | Continuously updated during Step 2 |
+| `02_fact_cards.md` *(splittable)* | Extracted facts and sources | Continuously updated during Step 3 |
 | `03_comparison_framework.md` | Selected framework and populated data | After Step 4 completion |
 | `04_reasoning_chain.md` | Fact → conclusion reasoning | After Step 6 completion |
 | `05_validation_log.md` | Use-case validation and review | After Step 7 completion |
-| `06_component_fit_matrix.md` | Exact-fit matrix for every proposed component/tool/pattern with status `Selected` / `Rejected` / `Experimental only` / `Needs user decision` | Before Step 8 deliverable formatting |
+| `06_component_fit_matrix.md` *(splittable)* | Exact-fit matrix for every proposed component/tool/pattern with status `Selected` / `Rejected` / `Experimental only` / `Needs user decision` | Before Step 8 deliverable formatting |
 | `OUTPUT_DIR/solution_draft##.md` | Complete solution draft | After Step 8 completion |
 | `OUTPUT_DIR/tech_stack.md` | Tech stack evaluation and decisions | After Phase 3 (optional) |
 | `OUTPUT_DIR/security_analysis.md` | Threat model and security controls | After Phase 4 (optional) |

.cursor/skills/research/steps/01_mode-a-initial-research.md

@@ -6,7 +6,9 @@ Triggered when no `solution_draft*.md` files exist in OUTPUT_DIR, or when the us
 
 **Role**: Professional software architect
 
-A focused preliminary research pass **before** the main solution research. The goal is to validate that the acceptance criteria and restrictions are realistic before designing a solution around them.
+> **AC must be design-independent**: describe testable outcomes only — no libraries, algorithms, params, or design choices. Implementation follows AC, never reverse. (IEEE 830 / Atlassian / GitScrum)
+
+A focused preliminary research pass **before** the main solution research. The goal is to validate that the acceptance criteria and restrictions are realistic before designing a solution around them. Any revision proposed in this phase must respect the design-independence rule above — propose AC changes as outcome/budget edits, not as implementation prescriptions.
 
 **Input**: All files from INPUT_DIR (or INPUT_FILE in standalone mode)
 
@@ -84,7 +86,7 @@ Full 8-step research methodology. Produces the first solution draft.
 
 Be concise in formulating. The fewer words, the better, but do not miss any important details.
 
-**Save action**: Write `RESEARCH_DIR/06_component_fit_matrix.md` before the final draft, then write `OUTPUT_DIR/solution_draft##.md` using template: `templates/solution_draft_mode_a.md`
+**Save action**: Write `RESEARCH_DIR/06_component_fit_matrix.md` (or its split-folder equivalent under `RESEARCH_DIR/06_component_fit_matrix/`, per the splittable-artifacts convention in `00_project-integration.md`) before the final draft, then write `OUTPUT_DIR/solution_draft##.md` using template: `templates/solution_draft_mode_a.md`
 
 ---
 

.cursor/skills/research/steps/02_mode-b-solution-assessment.md

@@ -29,6 +29,6 @@ Full 8-step research methodology applied to assessing and improving an existing
 9. For every revised candidate, prove exact fit against the Project Constraint Matrix. Do not select field-adjacent or "similar problem" options unless their intrinsic implementation constraints match the project.
 10. Based on findings, form a new solution draft in the same format
 
-**Save action**: Write `RESEARCH_DIR/06_component_fit_matrix.md` before the final draft, then write `OUTPUT_DIR/solution_draft##.md` (incremented) using template: `templates/solution_draft_mode_b.md`
+**Save action**: Write `RESEARCH_DIR/06_component_fit_matrix.md` (or its split-folder equivalent under `RESEARCH_DIR/06_component_fit_matrix/`, per the splittable-artifacts convention in `00_project-integration.md`) before the final draft, then write `OUTPUT_DIR/solution_draft##.md` (incremented) using template: `templates/solution_draft_mode_b.md`
 
 **Optional follow-up**: After Mode B completes, the user can request Phase 3 (Tech Stack Consolidation) or Phase 4 (Security Deep Dive) using the revised draft. These phases work identically to their Mode A descriptions in `steps/01_mode-a-initial-research.md`.

.cursor/skills/research/steps/03_engine-investigation.md

@@ -40,6 +40,7 @@ Key principle: Critical-sensitivity topics (AI/LLMs, blockchain) require sources
 - "What existing/competitor solutions address this problem?"
 - "What are the component parts of this problem?"
 - "For each component, what are the state-of-the-art solutions?"
+- "For each component, what are the practical alternatives across simple baseline, established production option, open-source option, commercial option, current SOTA, adjacent-domain option, and no-build/defer option?"
 - "What are the security considerations per component?"
 - "What are the cost implications of each approach?"
 
@@ -48,6 +49,7 @@ Key principle: Critical-sensitivity topics (AI/LLMs, blockchain) require sources
 - "What are the security vulnerabilities in the proposed architecture?"
 - "Where are the performance bottlenecks?"
 - "What solutions exist for each identified issue?"
+- "For each component already selected in the draft, what alternatives should be considered before keeping, replacing, or rejecting it?"
 
 **General sub-question patterns** (use when applicable):
 - **Sub-question A**: "What is X and how does it work?" (Definition & mechanism)
@@ -84,6 +86,27 @@ For **each sub-question**, generate **at least 3-5 search query variants** befor
 
 Record all planned queries in `00_question_decomposition.md` alongside each sub-question.
 
+#### Component Option Breadth (MANDATORY)
+
+Before Step 2, identify the component areas implied by the problem and create a search plan for options in each area. A component area is any replaceable tool, library, model, service, algorithm, data format, protocol, infrastructure pattern, or validation approach that could materially affect the solution.
+
+For every component area, generate search queries for these option families unless clearly not applicable:
+- **Simple baseline**: low-complexity classical or manual approach that can serve as a fallback or regression baseline.
+- **Established production option**: mature library/service/pattern with field usage.
+- **Open-source candidate**: permissive-license option with inspectable implementation and community history.
+- **Commercial/vendor option**: paid or vendor-supported option, including SDK/platform constraints.
+- **Current SOTA / research option**: recent model, paper, or benchmark leader that may be promising but immature.
+- **Adjacent-domain option**: solution from a neighboring domain with similar constraints.
+- **No-build / defer option**: whether the component can be avoided, simplified, or moved out of scope.
+- **Known bad option**: candidate or family that appears attractive but has documented failure modes or disqualifiers.
+
+For each component area, record:
+- Candidate names and option families to search.
+- At least 5 query variants covering alternatives, comparisons, limitations, licensing, runtime/scale, and exact project constraints.
+- The minimum evidence needed to mark a candidate `Selected`, `Rejected`, `Experimental only`, or `Needs user decision`.
+
+Add this as a "Component Option Search Plan" section in `00_question_decomposition.md`.
+
 **Research Subject Boundary Definition (BLOCKING - must be explicit)**:
 
 When decomposing questions, you must explicitly define the **boundaries of the research subject**:
@@ -123,6 +146,7 @@ Record the audit result in `00_question_decomposition.md` as a "Completeness Aud
    - List of decomposed sub-questions
    - **Chosen perspectives** (at least 3 from the Perspective Rotation table) with rationale
    - **Search query variants** for each sub-question (at least 3-5 per sub-question)
+   - **Component Option Search Plan** (component areas, option families, candidate names, query variants, required evidence)
    - **Completeness audit** (taxonomy cross-reference + domain discovery results)
 4. Write TodoWrite to track progress
 
@@ -136,7 +160,7 @@ Tier sources by authority, **prioritize primary sources** (L1 > L2 > L3 > L4). C
 
 **Tool Usage**:
 - Use `WebSearch` for broad searches; `WebFetch` to read specific pages
-- Use the `context7` MCP server (`resolve-library-id` then `get-library-docs`) for up-to-date library/framework documentation
+- Use the `context7` MCP server (`resolve-library-id` then `query-docs` / `get-library-docs`) for up-to-date library/framework documentation. **Mandatory per lead candidate** — see "API Capability Verification" below.
 - Always cross-verify training data claims against live sources for facts that may have changed (versions, APIs, deprecations, security advisories)
 - When citing web sources, include the URL and date accessed
 
@@ -149,6 +173,13 @@ Do not stop at the first few results. The goal is to build a comprehensive evide
 - Consult at least **2 different source tiers** per sub-question (e.g., L1 official docs + L4 community discussion)
 - If initial searches yield fewer than 3 relevant sources for a sub-question, **broaden the search** with alternative terms, related domains, or analogous problems
 
+**Minimum search effort per component area**:
+- Search every option family from the "Component Option Search Plan" before choosing a lead candidate.
+- For each lead, fallback, or rejected candidate, search at least one official/source-of-truth page and at least one independent validation source when available.
+- Search `"[component] alternatives"`, `"[candidate] vs [alternative]"`, `"[candidate] limitations"`, `"[candidate] license"`, `"[candidate] production"`, and `"[candidate] [binding project constraint]"`.
+- If fewer than 3 realistic candidates are found for a component area, explicitly document why the landscape is narrow and search adjacent domains before accepting that result.
+- Include at least one simple baseline and one "do not use" or disqualified candidate per component area when possible; these prevent false confidence in the selected option.
+
 **Candidate implementation-limit searches (MANDATORY)**:
 For every component/tool/library/service/pattern/algorithm that may be selected or recommended, search for its intrinsic implementation constraints. Do not rely on product category labels, marketing summaries, or examples from a different operating context. Include query variants for:
 - Official supported inputs/outputs, protocols, data formats, and deployment modes
@@ -159,6 +190,48 @@ For every component/tool/library/service/pattern/algorithm that may be selected
 - Licensing, security, maintenance, and community-health constraints
 - Exact phrases from the project's restrictions and acceptance criteria combined with the candidate name
 
+**API Capability Verification — Per-Mode (MANDATORY, BLOCKING for lead candidates)**:
+
+**Applicability**: this section applies only when the run is classified as **Technical-component selection** in the SKILL's Research Output Class section, and only to lead candidates that are libraries/SDKs/frameworks/services/protocols/data formats with multiple modes or configurations. For non-technical research (concept comparison, market/policy investigation, knowledge organization, root-cause analysis without tooling commitments), skip this entire sub-section and continue with the rest of Step 2 — the broader candidate implementation-limit search above is sufficient. State the skip explicitly once in `02_fact_cards.md` (or in `02_fact_cards/00_summary.md` if split): `API Capability Verification: not applicable — this run is a Non-technical investigation, no library/SDK/service candidates`.
+
+Most libraries/SDKs/services expose **multiple modes or configurations** (e.g., monocular vs stereo VO, sync vs async API, batch vs streaming inference, write-through vs write-behind cache). Selecting a candidate "because it supports X" without pinning *which mode* the project will use, and *whether that exact mode produces the required outputs from the required inputs*, is the most common silent-failure path in research. A library can support a class of problem in mode A while being unusable for the project's specific configuration in mode B.
+
+For every lead candidate that is a library/SDK/framework/service with multiple modes or configurations, do the following — in this order, before marking the candidate `Selected`:
+
+1. **Pin the exact mode/configuration the project will use.**
+   Derived from the Project Constraint Matrix: which inputs are available (sensor count, sensor types, data shapes, rates), which outputs are required (per `acceptance_criteria.md` and contract files), which hardware/runtime is fixed (per `restrictions.md`). Write this as a single sentence: "We will use `<library>` in `<mode/config>` with inputs `<list>` and expect outputs `<list>` on `<runtime>`." Do not progress past this step on a vague mode description.
+
+2. **Run `context7` (or equivalent docs lookup) for the candidate** — this is **mandatory for every lead library/SDK/framework candidate**, not optional. Minimum three queries per candidate:
+   1. *Mode enumeration*: "What modes/configurations does `<library>` support? List every value of the mode/config enum and what each requires as input."
+   2. *Project's exact mode*: "Show a minimum runnable example of `<library>` in `<the pinned mode>` with `<the project's input shape>`. What does it produce?"
+   3. *Disqualifier probe*: "Does `<library>` `<the pinned mode>` produce `<the required output>`? Are there published limitations of `<the pinned mode>` for `<the project's runtime/hardware>`?"
+
+   For services without context7 coverage, use official docs site + WebFetch on the API reference page + the project's example/tutorial directory in the source repo. Append every consulted URL to `01_source_registry.md` (or the appropriate category file under `01_source_registry/` if split — see splittable-artifacts convention in `00_project-integration.md`).
+
+3. **Save a Minimum Viable Example (MVE) for the pinned mode.**
+   Append to `02_fact_cards.md` / `02_fact_cards/` (or a sibling `02_mve_evidence.md`) at least one block per lead library candidate with:
+
+   ```markdown
+   ## MVE — <library> in <pinned mode>
+   - **Source**: <official URL or context7 reference, with date>
+   - **Inputs in the example**: <e.g., 2 calibrated cameras + IMU at 200 Hz>
+   - **Outputs in the example**: <e.g., 6-DoF pose with covariance>
+   - **Project inputs**: <e.g., 1 camera + IMU at 200 Hz>
+   - **Project outputs required**: <e.g., 6-DoF pose with metric translation>
+   - **Match assessment**: ✅ exact match / ⚠️ partial (specify dimension) / ❌ mismatch (specify dimension)
+   - **If ⚠️ or ❌**: cite the official-docs sentence that establishes the mismatch.
+   ```
+
+   If no official example covers the project's exact configuration → the candidate cannot be marked `Selected` based on category fit alone. Status must be `Experimental only` (with required-evidence note) or `Rejected` (when the docs explicitly disqualify the configuration).
+
+4. **Bind every numbered Restriction and Acceptance Criterion to the candidate's pinned mode.**
+   For each numbered line in `restrictions.md` and `acceptance_criteria.md`, decide one of: `Pass` (the pinned mode satisfies it with cited evidence), `Fail` (the pinned mode contradicts it with cited evidence), `Verify` (no evidence either way; deeper investigation required), `N/A` (the line is irrelevant to this component area). Record this in `02_fact_cards.md` (or the candidate's per-component file under `02_fact_cards/` if split) under the candidate's MVE block. The structural matrix in Step 7.5 reads from these bindings.
+
+5. **Treat "the same library in a different mode" as a different candidate.**
+   If the project's pinned mode is `Monocular` but the only documented evidence covers `Stereo`, do not silently soften "rotation only" into "rotation + translation". Open a separate candidate row for the Monocular mode, with its own MVE, fit assessment, and disqualifiers. Two modes of one library are two distinct candidates for the purposes of this gate.
+
+**Common silent-failure pattern this guards against**: a fact card paraphrases the docs as "supports A, B, C, D modes" when the docs actually mean "supports A; B; C and D as separate orthogonal modes". A category-level "Selected" decision then carries through every downstream artifact, masking that the project's required A+B combination does not exist as a single mode.
+
 **Search broadening strategies** (use when results are thin):
 - Try adjacent fields: if researching "drone indoor navigation", also search "robot indoor navigation", "warehouse AGV navigation"
 - Try different communities: academic papers, industry whitepapers, military/defense publications, hobbyist forums
@@ -170,7 +243,7 @@ For every component/tool/library/service/pattern/algorithm that may be selected
 **Search saturation rule**: Continue searching until new queries stop producing substantially new information. If the last 3 searches only repeat previously found facts, the sub-question is saturated.
 
 **Save action**:
-For each source consulted, **immediately** append to `01_source_registry.md` using the entry template from `references/source-tiering.md`.
+For each source consulted, **immediately** append to `01_source_registry.md` (or the appropriate category file under `01_source_registry/` if split) using the entry template from `references/source-tiering.md`.
 
 ---
 
@@ -200,7 +273,7 @@ Transform sources into **verifiable fact cards**:
   - ❓ Low: Inference or from unofficial sources
 
 **Save action**:
-For each extracted fact, **immediately** append to `02_fact_cards.md`:
+For each extracted fact, **immediately** append to `02_fact_cards.md` (or the appropriate category file under `02_fact_cards/` if split):
 ```markdown
 ## Fact #[number]
 - **Statement**: [specific fact description]
@@ -245,7 +318,7 @@ After initial fact extraction, review what you have found and identify **knowled
    - Failure cases and edge conditions
    - Recent developments that may change the picture
 
-4. **Update artifacts**: Append new sources to `01_source_registry.md`, new facts to `02_fact_cards.md`
+4. **Update artifacts**: Append new sources to `01_source_registry.md`, new facts to `02_fact_cards.md` (use the appropriate category files under `01_source_registry/` and `02_fact_cards/` if split)
 
 **Exit criteria**: Proceed to Step 4 when:
 - Every sub-question has at least 3 facts with at least one from L1/L2

.cursor/skills/research/steps/04_engine-analysis.md

@@ -26,6 +26,7 @@ Write to `03_comparison_framework.md`:
 
 **Required exact-fit dimensions for component/tool decisions**:
 When the output selects or recommends a component, tool, library, service, architecture pattern, or algorithm, the framework MUST include these dimensions unless explicitly not applicable:
+- Option family (`Simple baseline`, `Established production`, `Open-source`, `Commercial/vendor`, `Current SOTA`, `Adjacent-domain`, `No-build/defer`, `Known bad`)
 - Required inputs/outputs and ownership boundaries
 - Operating context and lifecycle fit
 - Non-functional envelope fit
@@ -33,6 +34,8 @@ When the output selects or recommends a component, tool, library, service, archi
 - Evidence quality and source tier
 - Selection status (`Selected`, `Rejected`, `Experimental only`, `Needs user decision`)
 
+For each component area, include multiple candidates in the initial population. Do not present only the preferred option unless the investigation found no realistic alternatives; if so, state the searches that proved the narrow landscape.
+
 ---
 
 ### Step 5: Reference Point Baseline Alignment
@@ -141,26 +144,61 @@ If using Y: [expected behavior]
 
 ### Step 7.5: Component Applicability Gate (BLOCKING)
 
-Before finalizing the solution draft, build an exact-fit matrix for every component/tool/library/service/pattern/algorithm that is selected, recommended, rejected, or treated as a fallback.
+**Applicability**: this gate applies only when the run is classified as **Technical-component selection** in the SKILL's Research Output Class section. For non-technical research (concept comparison, market/policy investigation, root-cause analysis without tooling, knowledge organization), skip this entire step and proceed to Step 8 — there are no components to gate. State the skip once in `05_validation_log.md`: `Step 7.5 (Component Applicability Gate): not applicable — Non-technical investigation`. For mixed runs (some component areas technical, some not), apply this gate only to the technical component areas; the non-technical ones do not produce 7.5 rows.
+
+Before finalizing the solution draft, build an exact-fit matrix for every component/tool/library/service/pattern/algorithm that is selected, recommended, rejected, or treated as a fallback. Free-form prose in a "Project Constraints Checked" column is **not sufficient** — mismatches hide inside rationale text. The matrix must be structured per restriction and per acceptance criterion.
+
+#### 7.5.1 Top-level Component Fit Matrix
 
 ```markdown
 # Component Fit Matrix
 
-| Candidate | Intended Role | Project Constraints Checked | Evidence | Mismatches / Disqualifiers | Status | Decision Rationale |
-|-----------|---------------|-----------------------------|----------|----------------------------|--------|--------------------|
-| [name] | [role] | [constraints] | [Fact # / Source #] | [none / list] | Selected / Rejected / Experimental only / Needs user decision | [why] |
+| Component Area | Candidate | Pinned Mode/Config | Option Family | Intended Role | API Capability Evidence | Mismatches / Disqualifiers | Status | Decision Rationale |
+|----------------|-----------|--------------------|---------------|---------------|-------------------------|----------------------------|--------|--------------------|
+| [area] | [name] | [exact mode/config the project will use, copied verbatim from the MVE block in Step 2] | [family] | [role] | MVE: [link to MVE block in `02_fact_cards.md` / `02_fact_cards/` or `02_mve_evidence.md`]; docs: [Source #] | [none / list] | Selected / Rejected / Experimental only / Needs user decision | [why] |
 ```
 
-Rules:
-- `Selected` is allowed only when the candidate's documented implementation assumptions match the project's explicit constraints and acceptance criteria.
-- `Experimental only` is required when a candidate might work but lacks proof for the exact operating context.
-- `Rejected` is required when documented assumptions conflict with project constraints.
-- `Needs user decision` is required when a mismatch changes scope, cost, safety, product behavior, or acceptance criteria.
+The new **Pinned Mode/Config** column is mandatory. A row without a pinned mode is incomplete. The new **API Capability Evidence** column links to the Minimum Viable Example saved during Step 2's API Capability Verification — without an MVE link the candidate cannot be `Selected`.
+
+#### 7.5.2 Restrictions × Candidate-Modes Sub-Matrix (MANDATORY)
+
+For each lead candidate row in the top-level matrix, append a structured cross-check that walks every numbered line of `restrictions.md` and `acceptance_criteria.md` against the candidate's **pinned mode/config**.
+
+```markdown
+## Sub-Matrix — <Candidate Name> in <Pinned Mode>
+
+| Restriction / AC | Candidate-mode behavior | Result | Evidence |
+|------------------|-------------------------|--------|----------|
+| R1: <verbatim line from restrictions.md> | <how the pinned mode behaves under this restriction> | ✅ Pass / ❌ Fail / ❓ Verify / N/A | [Fact # / Source # / MVE link] |
+| R2: ... | ... | ... | ... |
+| ... | ... | ... | ... |
+| AC-1.1: <verbatim line from acceptance_criteria.md> | <how the pinned mode satisfies (or contradicts) this AC's measurable target> | ✅ / ❌ / ❓ / N/A | [Fact # / Source # / MVE link] |
+| AC-1.2: ... | ... | ... | ... |
+| ... | ... | ... | ... |
+```
+
+Cell semantics:
+- ✅ **Pass** — the candidate's pinned mode satisfies this line, with cited official-doc or MVE evidence.
+- ❌ **Fail** — the candidate's pinned mode contradicts this line, with cited evidence. Even one ❌ disqualifies the candidate from `Selected` status.
+- ❓ **Verify** — no evidence yet either way; further investigation required (loops back to Step 2 / Step 3.5). A row left ❓ at the end of analysis blocks the candidate.
+- **N/A** — the line is irrelevant to this component area (state why in one phrase).
+
+A candidate row may not be marked `Selected` while any cell is ❌ or ❓.
+
+#### 7.5.3 Decision Rules
+
+- `Selected` is allowed only when (a) the top-level row has an MVE link, (b) the sub-matrix has zero ❌, (c) the sub-matrix has zero ❓, and (d) the candidate's documented implementation assumptions match the project's explicit constraints and acceptance criteria.
+- `Experimental only` is required when a candidate might work but lacks proof for the exact operating context (e.g., MVE exists for a similar configuration but not the exact one).
+- `Rejected` is required when documented assumptions conflict with project constraints (any sub-matrix row is ❌ with cited evidence).
+- `Needs user decision` is required when a mismatch changes scope, cost, safety, product behavior, or acceptance criteria — and the user has not yet been consulted.
+- Each component area must include at least one selected or fallback-safe option, plus the most credible rejected/experimental alternatives discovered during web research.
+- A component area with only one candidate is incomplete unless `00_question_decomposition.md` documents the broader searches and why they yielded no realistic alternatives.
 - A candidate may not appear as the lead solution in Step 8 unless this gate marks it `Selected`.
+- "Validation gate required" footnotes are not equivalent to `Selected`. If the validation gate concerns API capability (does the mode produce the required output?), that is a Step-2 / Step-7.5 question and must be resolved here, not deferred to runtime. Only validation gates concerning *runtime quality* (e.g., "does this VO converge on this terrain class?") may be carried forward as `Selected with runtime gate`.
 
-**Save action**: Write `06_component_fit_matrix.md`.
+**Save action**: Write `06_component_fit_matrix.md` (or, when split, the equivalent files under `06_component_fit_matrix/` — typically `00_summary.md` for the top-level matrix plus per-component sub-matrix files) containing both 7.5.1 (top-level) and 7.5.2 (per-candidate sub-matrices).
 
-**BLOCKING**: If any lead candidate is `Experimental only`, `Rejected`, or `Needs user decision`, do not silently proceed. Ask the user or choose a different selected candidate.
+**BLOCKING**: If any lead candidate has ❌, ❓, `Experimental only`, `Rejected`, or `Needs user decision` status, do not silently proceed. Ask the user or choose a different selected candidate.
 
 ---
 
@@ -175,8 +213,8 @@ Integrate all intermediate artifacts. Write to `OUTPUT_DIR/solution_draft##.md`
 
 Sources to integrate:
 - Extract background from `00_question_decomposition.md`
-- Reference key facts from `02_fact_cards.md`
+- Reference key facts from `02_fact_cards.md` (or files under `02_fact_cards/` if split)
 - Organize conclusions from `04_reasoning_chain.md`
-- Generate references from `01_source_registry.md`
+- Generate references from `01_source_registry.md` (or files under `01_source_registry/` if split)
 - Supplement with use cases from `05_validation_log.md`
 - For Mode A: include AC assessment from `00_ac_assessment.md`

.cursor/skills/research/templates/solution_draft_mode_a.md

@@ -10,17 +10,21 @@
 
 [Architecture solution that meets restrictions and acceptance criteria.]
 
+> **Applicability** — the table columns `Pinned Mode/Config` and `API Capability Evidence` apply only to technical-component runs (per SKILL.md → Research Output Class). For non-technical research outputs (concept comparison, market/policy report, investigation answer), this Architecture section may be replaced with a comparison/analysis section that does not use these columns; or the columns may be marked `N/A` per row when the row describes a non-technical "component" (a process, a policy, an organizational construct). For mixed runs, fill the columns only on rows that describe libraries/SDKs/frameworks/services/protocols/data formats/algorithms.
+
 ### Component: [Component Name]
 
-| Solution | Tools | Advantages | Limitations | Requirements | Security | Cost | Fit |
-|----------|-------|-----------|-------------|-------------|----------|------|-----|
-| [Option 1] | [lib/platform] | [pros] | [cons] | [intrinsic requirements] | [security] | [cost] | [Selected / Rejected / Experimental only / Needs user decision — cite exact-fit evidence and disqualifiers] |
-| [Option 2] | [lib/platform] | [pros] | [cons] | [intrinsic requirements] | [security] | [cost] | [Selected / Rejected / Experimental only / Needs user decision — cite exact-fit evidence and disqualifiers] |
+| Solution | Tools | Pinned Mode/Config | Advantages | Limitations | Requirements | Security | Cost | API Capability Evidence | Fit |
+|----------|-------|--------------------|-----------|-------------|-------------|----------|------|-------------------------|-----|
+| [Option 1] | [lib/platform] | [exact mode/config used: inputs, outputs, runtime] | [pros] | [cons] | [intrinsic requirements] | [security] | [cost] | MVE: [link to MVE block]; docs: [Source #] | [Selected / Rejected / Experimental only / Needs user decision — cite exact-fit evidence and disqualifiers] |
+| [Option 2] | [lib/platform] | [exact mode/config used] | [pros] | [cons] | [intrinsic requirements] | [security] | [cost] | MVE: [link]; docs: [Source #] | [Selected / Rejected / Experimental only / Needs user decision] |
 
 **Exact-fit evidence**:
 - Project constraints checked: [inputs/outputs, operating context, lifecycle, NFRs, acceptance criteria]
 - Evidence: [Fact # / Source #]
 - Disqualifiers: [none or list]
+- Restrictions × Candidate-Modes sub-matrix: see `06_component_fit_matrix.md` (or `06_component_fit_matrix/` if split) § <Candidate Name>
+- API capability gates: ✅ MVE saved / ⚠️ partial — see disqualifiers / ❌ no MVE — candidate is Experimental only or Rejected
 
 [Repeat per component]
 

.cursor/skills/research/templates/solution_draft_mode_b.md

@@ -13,17 +13,21 @@
 
 [Architecture solution that meets restrictions and acceptance criteria.]
 
+> **Applicability** — the table columns `Pinned Mode/Config` and `API Capability Evidence` apply only to technical-component runs (per SKILL.md → Research Output Class). For non-technical assessment outputs (e.g., reassessing a policy approach, comparing organizational designs), this Architecture section may be replaced with the assessment content that does not use these columns; or the columns may be marked `N/A` per row for non-technical "components". For mixed runs, fill the columns only on rows that describe libraries/SDKs/frameworks/services/protocols/data formats/algorithms.
+
 ### Component: [Component Name]
 
-| Solution | Tools | Advantages | Limitations | Requirements | Security | Performance | Fit |
-|----------|-------|-----------|-------------|-------------|----------|------------|-----|
-| [Option 1] | [lib/platform] | [pros] | [cons] | [intrinsic requirements] | [security] | [perf] | [Selected / Rejected / Experimental only / Needs user decision — cite exact-fit evidence and disqualifiers] |
-| [Option 2] | [lib/platform] | [pros] | [cons] | [intrinsic requirements] | [security] | [perf] | [Selected / Rejected / Experimental only / Needs user decision — cite exact-fit evidence and disqualifiers] |
+| Solution | Tools | Pinned Mode/Config | Advantages | Limitations | Requirements | Security | Performance | API Capability Evidence | Fit |
+|----------|-------|--------------------|-----------|-------------|-------------|----------|------------|-------------------------|-----|
+| [Option 1] | [lib/platform] | [exact mode/config used: inputs, outputs, runtime] | [pros] | [cons] | [intrinsic requirements] | [security] | [perf] | MVE: [link to MVE block]; docs: [Source #] | [Selected / Rejected / Experimental only / Needs user decision — cite exact-fit evidence and disqualifiers] |
+| [Option 2] | [lib/platform] | [exact mode/config used] | [pros] | [cons] | [intrinsic requirements] | [security] | [perf] | MVE: [link]; docs: [Source #] | [Selected / Rejected / Experimental only / Needs user decision] |
 
 **Exact-fit evidence**:
 - Project constraints checked: [inputs/outputs, operating context, lifecycle, NFRs, acceptance criteria]
 - Evidence: [Fact # / Source #]
 - Disqualifiers: [none or list]
+- Restrictions × Candidate-Modes sub-matrix: see `06_component_fit_matrix.md` (or `06_component_fit_matrix/` if split) § <Candidate Name>
+- API capability gates: ✅ MVE saved / ⚠️ partial — see disqualifiers / ❌ no MVE — candidate is Experimental only or Rejected
 
 [Repeat per component]
 

.cursor/skills/retrospective/SKILL.md

@@ -2,9 +2,9 @@
 name: retrospective
 description: |
   Collect metrics from implementation batch reports and code review findings, analyze trends across cycles,
-  and produce improvement reports with actionable recommendations.
-  3-step workflow: collect metrics, analyze trends, produce report.
-  Outputs to _docs/06_metrics/.
+  and produce improvement reports plus a lessons-log update with actionable recommendations.
+  4-step workflow: collect metrics, analyze trends, produce report, update lessons log.
+  Outputs to _docs/06_metrics/ and appends to _docs/LESSONS.md (ring buffer, last 15).
   Trigger phrases:
   - "retrospective", "retro", "run retro"
   - "metrics review", "feedback loop"
@@ -232,7 +232,7 @@ Present the report summary to the user.
 
 ```
 ┌────────────────────────────────────────────────────────────────┐
-│              Retrospective (3-Step Method)                     │
+│              Retrospective (4-Step Method)                     │
 ├────────────────────────────────────────────────────────────────┤
 │ PREREQ: batch reports exist in _docs/03_implementation/        │
 │                                                                │

.cursor/skills/test-run/SKILL.md

@@ -22,7 +22,7 @@ test-run has two modes. The caller passes the mode explicitly; if missing, defau
 | Mode | Scope | Typical caller | Input artifacts |
 |------|-------|---------------|-----------------|
 | `functional` (default) | Unit / integration / blackbox tests — correctness | autodev Steps that verify after Implement Tests or Implement | `scripts/run-tests.sh`, `_docs/02_document/tests/environment.md`, `_docs/02_document/tests/blackbox-tests.md` |
-| `perf` | Performance / load / stress / soak tests — latency, throughput, error-rate thresholds | autodev greenfield Step 9, existing-code Step 15 (pre-deploy) | `scripts/run-performance-tests.sh`, `_docs/02_document/tests/performance-tests.md`, AC thresholds in `_docs/00_problem/acceptance_criteria.md` |
+| `perf` | Performance / load / stress / soak tests — latency, throughput, error-rate thresholds | autodev greenfield Step 15, existing-code Step 15 (pre-deploy) | `scripts/run-performance-tests.sh`, `_docs/02_document/tests/performance-tests.md`, AC thresholds in `_docs/00_problem/acceptance_criteria.md` |
 
 Direct user invocation (`/test-run`) defaults to `functional`. If the user says "perf tests", "load test", "performance", or passes a performance scenarios file, run `perf` mode.
 
@@ -32,6 +32,17 @@ After selecting a mode, read its corresponding workflow below; do not mix them.
 
 ## Functional Mode
 
+### 0. System-Under-Test Reality Gate
+
+Before accepting any functional, blackbox, or e2e result as a pass, verify what the tests actually exercised.
+
+1. If `_docs/00_problem/input_data/expected_results/results_report.md` exists, at least one e2e/blackbox run must compare actual product outputs against that mapping or the machine-readable files it references.
+2. Stubs are allowed only for external systems outside the product boundary: flight controller/SITL, QGC observer, satellite-provider/Suite service, physical Jetson hardware, physical camera, unavailable licensed datasets, and network services.
+3. Stubs, fakes, deterministic fallbacks, monkeypatches, or direct replacement of internal product modules are not allowed for the behavior under test. Internal examples include VIO, safety/anchor wrapper, satellite retrieval, anchor verification, tile manager, MAVLink output adapter, FDR, and the A-Z localization pipeline.
+4. If tests pass only because an internal module is fake/scaffolded, classify the run as **failed** with category `missing product implementation`.
+5. If a scenario is blocked because external hardware/data is absent, verify the production code path exists before accepting the block as legitimate. Missing internal production code is not an environment block.
+6. If the test runner writes CSV/Markdown reports, inspect them. A zero exit code is not enough; blocked/internal-stubbed scenarios still require classification.
+
 ### 1. Detect Test Runner
 
 Check in order — first match wins:
@@ -94,7 +105,7 @@ Categorize skips as: **explicit skip (dead code)**, **runtime skip (unreachable)
 
 ### 5. Handle Outcome
 
-**All tests pass, zero skipped** → return success to the autodev for auto-chain.
+**All tests pass, zero skipped, and the System-Under-Test Reality Gate passes** → return success to the autodev for auto-chain.
 
 **Any test fails or errors** → this is a **blocking gate**. Never silently ignore failures. **Always investigate the root cause before deciding on an action.** Read the failing test code, read the error output, check service logs if applicable, and determine whether the bug is in the test or in the production code.
 

.cursor/skills/test-spec/SKILL.md

@@ -202,12 +202,12 @@ If invoked in `cycle-update` mode (see "Invocation Modes" above), read and follo
 | Missing acceptance_criteria.md, restrictions.md, or input_data/ | **STOP** — specification cannot proceed |
 | Missing input_data/expected_results/results_report.md | **STOP** — ask user to provide expected results mapping using the template |
 | Ambiguous requirements | ASK user |
-| Input data coverage below 75% (Phase 1) | Search internet for supplementary data, ASK user to validate |
+| Input data coverage below the canonical threshold (Phase 1) | Search internet for supplementary data, ASK user to validate. See `.cursor/rules/cursor-meta.mdc` Quality Thresholds for the canonical 75% number — do not hardcode a different threshold here. |
 | Expected results missing or not quantifiable (Phase 1) | ASK user to provide quantifiable expected results before proceeding |
 | Test scenario conflicts with restrictions | ASK user to clarify intent |
 | System interfaces unclear (no architecture.md) | ASK user or derive from solution.md |
 | Test data or expected result not provided for a test scenario (Phase 3) | WARN user and REMOVE the test |
-| Final coverage below 75% after removals (Phase 3) | BLOCK — require user to supply data or accept reduced spec |
+| Final coverage below the canonical threshold after removals (Phase 3) | BLOCK — require user to supply data or accept reduced spec (see `cursor-meta.mdc` Quality Thresholds) |
 
 ## Common Mistakes
 
@@ -252,7 +252,8 @@ When the user wants to:
 │                                                                      │
 │ Phase 3: Test Data & Expected Results Validation Gate (HARD GATE)    │
 │   → phases/03-data-validation-gate.md                                │
-│   [BLOCKING: coverage ≥ 75% required to pass]                        │
+│   [BLOCKING: coverage ≥ canonical threshold required to pass —      │
+│    see cursor-meta.mdc Quality Thresholds (75%)]                    │
 │                                                                      │
 │ Hardware-Dependency Assessment (BLOCKING, pre-Phase-4)               │
 │   → phases/hardware-assessment.md                                    │

.cursor/skills/test-spec/phases/03-data-validation-gate.md

@@ -1,7 +1,7 @@
 # Phase 3: Test Data & Expected Results Validation Gate (HARD GATE)
 
 **Role**: Professional Quality Assurance Engineer
-**Goal**: Ensure every test scenario produced in Phase 2 has concrete, sufficient test data. Remove tests that lack data. Verify final coverage stays above 75%.
+**Goal**: Ensure every test scenario produced in Phase 2 has concrete, sufficient test data. Remove tests that lack data. Verify final coverage stays above the canonical threshold (currently 75% — see `.cursor/rules/cursor-meta.mdc` Quality Thresholds; never hardcode a different number in any phase).
 **Constraints**: This phase is MANDATORY and cannot be skipped.
 
 ## Step 1 — Build the requirements checklist

.cursor/skills/test-spec/templates/expected-results.md

@@ -95,7 +95,7 @@ Examples:
 
 File: `expected_results/image_01_detections.json`
 
-​```json
+```json
 {
   "input": "image_01.jpg",
   "expected": {
@@ -119,7 +119,7 @@ File: `expected_results/image_01_detections.json`
     ]
   }
 }
-​```
+```
 ```
 
 ---

.dockerignore

@@ -0,0 +1,29 @@
+.git/
+.github/
+.venv/
+venv/
+env/
+__pycache__/
+*.py[cod]
+.pytest_cache/
+.mypy_cache/
+.ruff_cache/
+.coverage*
+htmlcov/
+build/
+dist/
+_skbuild/
+CMakeFiles/
+CMakeCache.txt
+cmake_install.cmake
+*.engine
+*.calib
+*.index
+*.faiss
+*.onnx
+tests/fixtures/large_replays/
+tests/fixtures/flight_derkachi/
+tests/fixtures/tiles_corpus/
+_docs/
+*.log
+.DS_Store

.editorconfig

@@ -0,0 +1,24 @@
+root = true
+
+[*]
+charset = utf-8
+end_of_line = lf
+insert_final_newline = true
+trim_trailing_whitespace = true
+indent_style = space
+indent_size = 4
+
+[*.{yml,yaml,json,toml}]
+indent_size = 2
+
+[*.{cpp,c,h,hpp,cc,hh}]
+indent_size = 4
+
+[*.{cmake,CMakeLists.txt}]
+indent_size = 2
+
+[Makefile]
+indent_style = tab
+
+[*.md]
+trim_trailing_whitespace = false

.env.example

@@ -0,0 +1,53 @@
+# gps-denied-onboard — environment variables
+# See _docs/02_document/module-layout.md and AZ-263_initial_structure.md § Environment Variables.
+
+# Required: selects the FC adapter at the composition root.
+# One of: ardupilot_plane | inav
+GPS_DENIED_FC_PROFILE=ardupilot_plane
+
+# Required: runtime tier gate; 1=workstation/CI, 2=Jetson production
+GPS_DENIED_TIER=1
+
+# Required: Postgres connection used by C6 (tile cache + descriptor index)
+DB_URL=postgresql://gps_denied:dev@db:5432/gps_denied
+
+# Required (dev/operator only): satellite-provider base URL for tile download
+# Not set in flight (no egress)
+SATELLITE_PROVIDER_URL=http://mock-sat:5100
+
+# Required: path to JSON camera calibration loaded at startup
+CAMERA_CALIBRATION_PATH=/fixtures/calibration/adti26.json
+
+# Required: structured log level (DEBUG | INFO | WARNING | ERROR)
+LOG_LEVEL=DEBUG
+
+# Required: structured log sink (console | journald | fdr)
+LOG_SINK=console
+
+# Required (production): per-flight MAVLink 2.0 signing key path
+# Dev key from tests/fixtures/mavlink_signing/dev_key in dev-tier1.
+MAVLINK_SIGNING_KEY=tests/fixtures/mavlink_signing/dev_key
+
+# CMake / runtime BUILD_* gating flags
+# Defaults below match the airborne deployment binary (ADR-002 / ADR-011).
+# Strategy flags use OFF for opt-in non-default strategies; ON for the
+# deployment defaults that the runtime expects to be linked.
+BUILD_VINS_MONO=OFF
+BUILD_SALAD=OFF
+BUILD_C11_TILE_MANAGER=OFF
+# Replay-mode strategy flags (ADR-011) — must be ON in the airborne and
+# research binaries so replay can run from the same image. The CI test
+# compose files already set these explicitly; production sets them ON.
+# BUILD_VIDEO_FILE_FRAME_SOURCE=ON
+# BUILD_TLOG_REPLAY_ADAPTER=ON
+# BUILD_REPLAY_SINK_JSONL=ON
+# Dev-only: enables `signing_key_source='dev_static'` on the AP FC adapter.
+# MUST stay OFF on production images; ON only in dev/CI containers.
+# BUILD_DEV_STATIC_KEY=OFF
+
+# Required: C7 inference backend (tensorrt | pytorch_fp16 | onnx_trt_ep)
+INFERENCE_BACKEND=pytorch_fp16
+
+# Required: filesystem paths for runtime artifacts
+FDR_PATH=/var/lib/gps-denied/fdr
+TILE_CACHE_PATH=/var/lib/gps-denied/tiles

.env.test.example

@@ -0,0 +1,42 @@
+# AZ-688: dev-only environment for the Jetson e2e harness.
+# Jetson-only test policy (2026-05-20) — see _docs/LESSONS.md.
+#
+# Copy this file to `.env.test` and customize. NEVER commit `.env.test`
+# (gitignored). Sourced by `scripts/run-tests-jetson.sh` before
+# `docker compose up`.
+
+# Suite JWT contract — see ../_docs/10_auth.md. The same secret signs the
+# dev JWT (AZ-690) and validates it at the satellite-provider boundary.
+# MUST be ≥ 32 bytes UTF-8. Generate a fresh value with:
+#   openssl rand -hex 32
+JWT_SECRET=DEV-ONLY-REPLACE-WITH-OPENSSL-RAND-HEX-32-OUTPUT-XXXXXXX
+
+# JWT issuer / audience claims. Dev-only values that ONLY validate against
+# the dev secret above. Production deploys MUST use real values provided
+# by the admin team (the admin API stamps `iss`; satellite-provider
+# validates `aud`).
+JWT_ISSUER=DEV-ONLY-iss-admin-azaion-local
+JWT_AUDIENCE=DEV-ONLY-aud-satellite-provider
+
+# Google Maps Platform key. Left empty: AZ-689 seeds local fixture tiles
+# instead, so the hermetic Derkachi e2e flow never calls GoogleMaps. If
+# you need to exercise the real GMaps tile-download path, set this to a
+# valid key.
+GOOGLE_MAPS_API_KEY=
+
+# AZ-777: Bearer token C11 sends to satellite-provider as
+# `Authorization: Bearer <token>`. The token is a JWT signed with
+# JWT_SECRET above and stamped with the same iss/aud the provider
+# validates. Mint a dev token with:
+#   python scripts/mint_dev_jwt.py
+# Production deploys retrieve this from the admin API and rotate per
+# operator session — never commit a real one.
+SATELLITE_PROVIDER_API_KEY=PASTE-MINTED-JWT-HERE
+
+# SECURITY: development-only TLS bypass for the parent-suite
+# satellite-provider self-signed dev cert. The compose env block sets
+# SATELLITE_PROVIDER_TLS_INSECURE=1 — it stays inside the Jetson e2e
+# harness, never in production. Production deploys MUST use a real
+# CA-issued cert (or your own internal CA) and leave this unset (or
+# set to "0"). C11 logs a single WARNING at startup whenever the
+# insecure flag is active so the operator can audit it.

.gitattributes

@@ -0,0 +1,4 @@
+_docs/00_problem/input_data/flight_derkachi/flight_derkachi.mp4 filter=lfs diff=lfs merge=lfs -text
+models/**/*.pt filter=lfs diff=lfs merge=lfs -text
+models/**/*.onnx filter=lfs diff=lfs merge=lfs -text
+models/**/*.engine filter=lfs diff=lfs merge=lfs -text

.github/pull_request_template.md

@@ -1,15 +0,0 @@
-## Summary
-[1-3 bullet points describing the change]
-
-## Related Tasks
-[JIRA-ID links]
-
-## Testing
-- [ ] Unit tests pass
-- [ ] Integration tests pass
-- [ ] Manual testing done (if applicable)
-
-## Checklist
-- [ ] No new linter warnings
-- [ ] No secrets committed
-- [ ] API docs updated (if applicable)

.github/workflows/ci-tier2.yml

@@ -0,0 +1,25 @@
+name: ci-tier2
+
+on:
+  push:
+    branches: [stage, main]
+  workflow_dispatch:
+
+jobs:
+  build-tier2:
+    runs-on: [self-hosted, jetson, orin-nano-super]
+    steps:
+      - uses: actions/checkout@v4
+      - name: Native build (deployment)
+        run: |
+          cmake -S . -B build -DBUILD_VINS_MONO=OFF -DBUILD_VPR_SALAD=OFF -DBUILD_C11_TILE_MANAGER=OFF
+          cmake --build build --parallel
+
+  ac-bound-nfts:
+    runs-on: [self-hosted, jetson, orin-nano-super]
+    needs: build-tier2
+    steps:
+      - uses: actions/checkout@v4
+      - name: AC-bound NFTs (NFT-PERF / NFT-LIM / NFT-RES / NFT-SEC / IT-12)
+        run: |
+          pytest -m tier2 -q tests/perf tests/security tests/resilience

.github/workflows/ci.yml

@@ -0,0 +1,104 @@
+name: ci-tier1
+
+on:
+  push:
+    branches: [dev, stage, main]
+  pull_request:
+    branches: [dev, stage, main]
+
+jobs:
+  lint:
+    runs-on: ubuntu-22.04
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+      # AZ-300 — `[inference]` (torch + torchvision + onnxruntime) is now
+      # required for `mypy src` to type-check `c7_inference.pytorch_fp16_runtime`
+      # and for `pytest` to collect `test_pytorch_fp16_runtime.py`. Tier-1
+      # CI uses the CPU-only torch wheel; CUDA-gated tests skip themselves
+      # via `pytest.mark.skipif(not torch.cuda.is_available(), ...)`.
+      - run: pip install -e ".[dev,inference]"
+      - run: ruff check src tests
+      - run: mypy src
+
+  unit:
+    runs-on: ubuntu-22.04
+    needs: lint
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+      - run: pip install -e ".[dev,inference]"
+      - name: pytest unit (per-component coverage gate)
+        run: pytest -q --cov=gps_denied_onboard --cov-fail-under=75 tests/unit
+
+  integration:
+    runs-on: ubuntu-22.04
+    needs: unit
+    steps:
+      - uses: actions/checkout@v4
+      - name: docker compose up
+        run: docker compose -f docker-compose.test.yml up --abort-on-container-exit --exit-code-from e2e-runner --build
+
+  build:
+    name: build-${{ matrix.kind }}
+    runs-on: ubuntu-22.04
+    needs: lint
+    strategy:
+      fail-fast: false
+      matrix:
+        kind: [deployment, research]
+        include:
+          # AZ-332 — BUILD_OKVIS2 forced OFF in Tier-1 CI until the tier2
+          # follow-up wires `okvis::ThreadedKFVio` end-to-end. The C++
+          # binding skeleton + CMake glue still ship in this build; full
+          # OKVIS2 native compile is gated on installing Ceres-solver +
+          # OKVIS2 vendored submodules (BRISK, DBoW2) via apt, plus
+          # `submodules: recursive` checkout. That CI lift is the
+          # tier2 task's surface, not AZ-332's.
+          - kind: deployment
+            cmake_flags: >-
+              -DBUILD_OKVIS2=OFF -DBUILD_VINS_MONO=OFF
+              -DBUILD_VPR_SALAD=OFF -DBUILD_C11_TILE_MANAGER=OFF
+          - kind: research
+            cmake_flags: >-
+              -DBUILD_OKVIS2=OFF -DBUILD_VINS_MONO=ON -DBUILD_VPR_SALAD=ON
+    steps:
+      - uses: actions/checkout@v4
+      - run: cmake -S . -B build ${{ matrix.cmake_flags }}
+      - run: cmake --build build --parallel
+
+  sbom-diff:
+    runs-on: ubuntu-22.04
+    needs: build
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+      - name: SBOM diff (ADR-002 enforcement)
+        run: python ci/sbom_diff.py --deployment build-deployment-sbom.json --research build-research-sbom.json
+
+  security:
+    runs-on: ubuntu-22.04
+    needs: build
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+      - run: pip install pip-audit
+      - run: pip-audit -r pyproject.toml || true
+      - name: OpenCV pin gate (D-CROSS-CVE-1)
+        run: python ci/opencv_pin_gate.py --pyproject pyproject.toml
+
+  push-images:
+    runs-on: ubuntu-22.04
+    if: github.event_name == 'push' && contains(fromJson('["refs/heads/dev","refs/heads/stage","refs/heads/main"]'), github.ref)
+    needs: [unit, integration, build, sbom-diff, security]
+    steps:
+      - uses: actions/checkout@v4
+      - run: echo "push images to GHCR (deployment + research) — wiring lands per release task"

.github/workflows/cve-rescan.yml

@@ -0,0 +1,19 @@
+name: cve-rescan
+
+on:
+  schedule:
+    - cron: "0 5 1 * *"   # 05:00 UTC on the 1st of each month
+  workflow_dispatch:
+
+jobs:
+  rescan:
+    runs-on: ubuntu-22.04
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+      - run: pip install pip-audit
+      - run: pip-audit -r pyproject.toml
+      - name: OpenCV pin gate (D-CROSS-CVE-1)
+        run: python ci/opencv_pin_gate.py --pyproject pyproject.toml

.github/workflows/release.yml

@@ -0,0 +1,24 @@
+name: release
+
+on:
+  push:
+    tags:
+      - "v*"
+
+jobs:
+  jetpack-image:
+    runs-on: [self-hosted, jetson, orin-nano-super]
+    steps:
+      - uses: actions/checkout@v4
+      - name: Build JetPack image
+        run: echo "JetPack image build + sign + attest — concrete wiring lands per deploy task"
+
+  operator-orchestrator-tarball:
+    runs-on: ubuntu-22.04
+    needs: jetpack-image
+    steps:
+      - uses: actions/checkout@v4
+      - name: Bundle operator-orchestrator tarball
+        run: |
+          mkdir -p dist
+          tar -czf dist/operator-orchestrator.tar.gz docker-compose.yml docker/ _docs/

.gitignore

@@ -1 +1,87 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+*.egg
+*.egg-info/
+.eggs/
+.pytest_cache/
+.coverage
+.coverage.*
+coverage.xml
+htmlcov/
+.mypy_cache/
+.ruff_cache/
+.tox/
+.venv/
+venv/
+env/
+
+# Build artifacts
+build/
+dist/
+_skbuild/
+CMakeFiles/
+CMakeCache.txt
+cmake_install.cmake
+Makefile
+compile_commands.json
+
+# Native engines and caches
+*.engine
+*.calib
+*.index
+*.faiss
+*.onnx
+*.trt
+
+# Test fixtures — large blobs are out-of-band
+tests/fixtures/large_replays/
+tests/fixtures/flight_derkachi/*.mp4
+tests/fixtures/flight_derkachi/*.h264
+tests/fixtures/flight_derkachi/*.tlog
+tests/fixtures/tiles_corpus/*.jpg
+tests/fixtures/tiles_corpus/*.png
+e2e/fixtures/sitl_replay/
+
+# Problem-folder flight-log inputs (binary, out-of-band)
+_docs/00_problem/input_data/**/*.tlog
+_docs/00_problem/input_data/**/*.mp4
+_docs/00_problem/input_data/**/*.h264
+_docs/00_problem/input_data/**/*.mkv
+_docs/00_problem/input_data/**/*.zip
+
+# Locally-generated evidence frames for extraction fixtures (large, regenerable)
+_docs/00_problem/input_data/**/frames_src/
+_docs/00_problem/input_data/**/frames_optA/
+_docs/00_problem/input_data/**/frames_optB/
+_docs/00_problem/input_data/**/frames_optC/
+
+# Editor / OS noise
+.idea/
+.vscode/
 .DS_Store
+Thumbs.db
+*.swp
+*~
+
+# Logs and runtime data
+*.log
+/var/lib/gps-denied/
+fdr_output/
+tile_cache/
+e2e-results/
+
+# Local scratch / one-off diagnostics
+_scratch/
+
+# Secrets
+.env
+.env.local
+.env.test
+*.key
+!tests/fixtures/mavlink_signing/dev_key
+
+# Deploy rollback bookmark (written by scripts/stop-services.sh)
+.previous-tags.env

.gitmodules

@@ -0,0 +1,6 @@
+[submodule "cpp/pybind11/upstream"]
+	path = cpp/pybind11/upstream
+	url = https://github.com/pybind/pybind11.git
+[submodule "cpp/okvis2/upstream"]
+	path = cpp/okvis2/upstream
+	url = https://github.com/smartroboticslab/okvis2.git

.woodpecker/01-test.yml

@@ -0,0 +1,43 @@
+# Cycle-1 trigger: manual-only.
+#
+# Rationale (per _docs/04_deploy/ci_cd_pipeline.md → Decision Record):
+#   The Tier-1 e2e harness (docker-compose.test.yml + tests/e2e/Dockerfile)
+#   is heavy: TensorRT-class pytorch fp16, gtsam, Postgres 16, and the
+#   Derkachi replay clip. It is shipped opt-in until per-run wall-clock on
+#   the colocated arm64 Jetson agent is characterised.
+#
+# Flip-back (cycle-2 polish item #1 in _docs/04_deploy/ci_cd_pipeline.md):
+#   1. Replace `event: [manual]` with `event: [push, pull_request, manual]`
+#      below.
+#   2. Add `depends_on: [01-test]` to .woodpecker/02-build-push.yml.
+
+when:
+  event: [manual]
+  branch: [dev, stage, main]
+
+matrix:
+  include:
+    - PLATFORM: arm64
+      TAG_SUFFIX: arm
+    # - PLATFORM: amd64
+    #   TAG_SUFFIX: amd
+
+labels:
+  platform: ${PLATFORM}
+
+steps:
+  - name: e2e
+    image: docker
+    commands:
+      - docker compose -f docker-compose.test.yml up --build --abort-on-container-exit --exit-code-from e2e-runner
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock
+
+  - name: down
+    image: docker
+    when:
+      status: [success, failure]
+    commands:
+      - docker compose -f docker-compose.test.yml down -v
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock

.woodpecker/02-build-push.yml

@@ -0,0 +1,85 @@
+# Cycle-1 trigger: push + manual on dev/stage/main, NO depends_on.
+#
+# Rationale (per _docs/04_deploy/ci_cd_pipeline.md → Decision Record):
+#   01-test.yml runs `event: [manual]` only in cycle-1, so a `depends_on:
+#   [01-test]` clause here would skip every push (no preceding test run to
+#   succeed against). The un-gated stance mirrors the `detections` deferral
+#   pattern documented in `../_infra/ci/README.md` → "detections deferral".
+#
+# Re-gate (cycle-2 polish item #1 in _docs/04_deploy/ci_cd_pipeline.md):
+#   Add `depends_on: [01-test]` below once .woodpecker/01-test.yml flips to
+#   `event: [push, pull_request, manual]`.
+#
+# Images pushed in cycle-1:
+#   - azaion/gps-denied-onboard-companion-tier1:${BRANCH}-${TAG_SUFFIX}
+#   - azaion/gps-denied-onboard-operator-orchestrator:${BRANCH}-${TAG_SUFFIX}
+#
+# Image NOT pushed in cycle-1 (reserved for cycle-2 / companion-jetson):
+#   - azaion/gps-denied-onboard:${BRANCH}-${TAG_SUFFIX}
+#     (parent-suite Jetson compose at ../_infra/deploy/jetson/docker-compose.yml
+#      expects this exact tag; cycle-1 must not write to it or Watchtower
+#      on fielded Jetsons will pull a Tier-1 dev image.)
+#
+# OCI labels (suite-mandated, AZ-204 — see ../_infra/ci/README.md → "OCI
+# image labels and commit provenance"):
+#   org.opencontainers.image.revision = $CI_COMMIT_SHA
+#   org.opencontainers.image.created  = <UTC RFC 3339>
+#   org.opencontainers.image.source   = $CI_REPO_URL
+# Plus --build-arg CI_COMMIT_SHA so the Dockerfile can bake ENV AZAION_REVISION.
+
+when:
+  event: [push, manual]
+  branch: [dev, stage, main]
+
+matrix:
+  include:
+    - PLATFORM: arm64
+      TAG_SUFFIX: arm
+    # - PLATFORM: amd64
+    #   TAG_SUFFIX: amd
+
+labels:
+  platform: ${PLATFORM}
+
+steps:
+  - name: build-push-companion-tier1
+    image: docker
+    environment:
+      REGISTRY_HOST:  { from_secret: registry_host }
+      REGISTRY_USER:  { from_secret: registry_user }
+      REGISTRY_TOKEN: { from_secret: registry_token }
+    commands:
+      - echo "$REGISTRY_TOKEN" | docker login "$REGISTRY_HOST" -u "$REGISTRY_USER" --password-stdin
+      - export TAG=${CI_COMMIT_BRANCH}-${TAG_SUFFIX}
+      - export BUILD_DATE=$(date -u +%Y-%m-%dT%H:%M:%SZ)
+      - |
+        docker build -f docker/companion-tier1.Dockerfile \
+          --build-arg CI_COMMIT_SHA=$CI_COMMIT_SHA \
+          --label org.opencontainers.image.revision=$CI_COMMIT_SHA \
+          --label org.opencontainers.image.created=$BUILD_DATE \
+          --label org.opencontainers.image.source=$CI_REPO_URL \
+          -t $REGISTRY_HOST/azaion/gps-denied-onboard-companion-tier1:$TAG .
+      - docker push $REGISTRY_HOST/azaion/gps-denied-onboard-companion-tier1:$TAG
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock
+
+  - name: build-push-operator-orchestrator
+    image: docker
+    environment:
+      REGISTRY_HOST:  { from_secret: registry_host }
+      REGISTRY_USER:  { from_secret: registry_user }
+      REGISTRY_TOKEN: { from_secret: registry_token }
+    commands:
+      - echo "$REGISTRY_TOKEN" | docker login "$REGISTRY_HOST" -u "$REGISTRY_USER" --password-stdin
+      - export TAG=${CI_COMMIT_BRANCH}-${TAG_SUFFIX}
+      - export BUILD_DATE=$(date -u +%Y-%m-%dT%H:%M:%SZ)
+      - |
+        docker build -f docker/operator-orchestrator.Dockerfile \
+          --build-arg CI_COMMIT_SHA=$CI_COMMIT_SHA \
+          --label org.opencontainers.image.revision=$CI_COMMIT_SHA \
+          --label org.opencontainers.image.created=$BUILD_DATE \
+          --label org.opencontainers.image.source=$CI_REPO_URL \
+          -t $REGISTRY_HOST/azaion/gps-denied-onboard-operator-orchestrator:$TAG .
+      - docker push $REGISTRY_HOST/azaion/gps-denied-onboard-operator-orchestrator:$TAG
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock

CMakeLists.txt

@@ -0,0 +1,32 @@
+cmake_minimum_required(VERSION 3.22)
+project(gps_denied_onboard LANGUAGES CXX)
+
+# Compile options ----------------------------------------------------------
+
+set(CMAKE_CXX_STANDARD 17)
+set(CMAKE_CXX_STANDARD_REQUIRED ON)
+set(CMAKE_CXX_EXTENSIONS OFF)
+set(CMAKE_POSITION_INDEPENDENT_CODE ON)
+
+if(NOT CMAKE_BUILD_TYPE)
+  set(CMAKE_BUILD_TYPE RelWithDebInfo CACHE STRING "Build type" FORCE)
+endif()
+
+# Helper modules -----------------------------------------------------------
+
+list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake")
+include(build_options)
+include(dependencies)
+include(strategies)
+
+# Native subprojects -------------------------------------------------------
+
+add_subdirectory(cpp)
+
+# Tests --------------------------------------------------------------------
+
+option(BUILD_TESTING "Enable native unit tests (C++ gtest)" OFF)
+if(BUILD_TESTING)
+  enable_testing()
+  add_subdirectory(cpp/tests)
+endif()

README.md

@@ -0,0 +1,26 @@
+# gps-denied-onboard
+
+Companion onboard system for GPS-denied UAV navigation. Detailed design and architecture documentation lives under [`_docs/`](_docs/).
+
+## Quick links
+
+- Problem statement: [`_docs/00_problem/problem.md`](_docs/00_problem/problem.md)
+- Architecture: [`_docs/02_document/architecture.md`](_docs/02_document/architecture.md)
+- Module layout (file ownership): [`_docs/02_document/module-layout.md`](_docs/02_document/module-layout.md)
+- Component docs: [`_docs/02_document/components/`](_docs/02_document/components/)
+- Test specs: [`_docs/02_document/tests/`](_docs/02_document/tests/)
+- Deployment: [`_docs/02_document/deployment/`](_docs/02_document/deployment/)
+
+## Local development
+
+```bash
+python -m venv .venv && source .venv/bin/activate
+pip install -e ".[dev]"
+pytest -q tests/unit/
+```
+
+For full Tier-1 integration via Docker, see [`_docs/02_document/deployment/containerization.md`](_docs/02_document/deployment/containerization.md).
+
+## Build matrix
+
+Four binaries built from this codebase: **airborne**, **research**, **operator-orchestrator**, **replay-cli**. CMake `BUILD_*` flags gate component inclusion per binary — see [`cmake/build_options.cmake`](cmake/build_options.cmake) and [`_docs/02_document/module-layout.md` § Build-Time Exclusion Map](_docs/02_document/module-layout.md#build-time-exclusion-map-adr-002).

_docs/00_problem/acceptance_criteria.md

@@ -1,155 +1,109 @@
 # Acceptance Criteria
 
-> **Last revised**: 2026-04-26 (post Mode B Solution Assessment + user-driven addendum on VPR granularity & change-robustness + user lock-in of Mode B open items Q1–Q5).
-> Changes vs. previous version (2026-04-25): AC-1.2 split into hard-floor + stretch; AC-1.4 made quantitative; AC-2.2 split per pipeline stage; AC-3.4 dual-trigger; AC-4.3 autopilot-pinned; AC-5.2 N pinned; AC-7.1 scoped to level flight; AC-8.2 freshness by sector; six new AC added (AC-NEW-1 … AC-NEW-6).
-> Changes 2026-04-26: AC-4.3 extended to dual-channel hybrid (GPS_INPUT primary + ODOMETRY auxiliary); AC-8.6 added (VPR retrieval-unit + change-robustness); AC-NEW-7 added with confirmed numeric thresholds (cache-poisoning safety budget).
+> Last revised 2026-05-07 (cleanup pass: stripped algorithm/library/parameter implementation details; renamed source label `vo_extrapolated` → `visual_propagated`; broadened FC scope to ArduPilot + iNav).
+> Subsequent revision 2026-05-07 (post-SQ6 research): AC-4.3 reworded to acknowledge that no single message type is accepted by both ArduPilot Plane and iNav — per-FC interface is named explicitly (MAVLink `GPS_INPUT` for ArduPilot Plane, MSP2 `MSP2_SENSOR_GPS` for iNav). Rationale and L1 sources in `_docs/00_research/02_fact_cards/SQ6_fc_external_positioning.md` / `_docs/00_research/01_source_registry/SQ6_external_positioning.md` Sources #4, #9, #10, #12, #13.
+> Subsequent revision 2026-05-09 (Plan Phase 2a.0 outcomes): AC-NEW-4 and AC-NEW-7 validation requirements relaxed from "≥100 flights" literal to Monte-Carlo-with-stated-CI over currently-available data corpus; multi-flight statistical headroom moved to Step 4 risk register (D-PROJ-3). AC-8.4 augmented with explicit in-air-no-upload security gate (flight-state process-level isolation; post-landing upload tool); local mid-flight tile format pinned to match `satellite-provider`'s on-disk format. AC-NEW-7 external-dependency note revised: parent-suite voting layer is not currently implemented; tracked as parent-suite design task D-PROJ-2.
+> See git history for prior versions.
 
 ## Position Accuracy
-
-- **AC-1.1** — The system shall determine GPS coordinates of frame centers within **50 m** of true GPS for **≥80%** of photos in normal flight segments.
-- **AC-1.2** — The system shall determine GPS coordinates of frame centers within **20 m** of true GPS for **≥50%** of photos in normal flight segments.
-- **AC-1.3** — Maximum cumulative VO drift between two consecutive satellite-anchored fixes shall be **<100 m** (VO-only fallback) or **<50 m** (when IMU is fused). Drift is measured as ‖VO-extrapolated centre − next anchor centre‖ at the moment of the anchor fix.
-- **AC-1.4** — The system shall report a **quantitative confidence score** per position estimate, comprising:
-  - the 95% covariance ellipse semi-major axis in meters, AND
-  - a categorical label `{satellite_anchored, vo_extrapolated, dead_reckoned}`.
+- **AC-1.1** — Frame-center GPS within **50 m** of true GPS for **≥80%** of normal-flight photos.
+- **AC-1.2** — Frame-center GPS within **20 m** of true GPS for **≥50%** of normal-flight photos.
+- **AC-1.3** — Cumulative drift between two consecutive satellite-anchored fixes: **<100 m** visual-only / **<50 m** with IMU fused. Measured as ‖propagated centre − next anchor centre‖ at anchor fix. Every estimate carries `last_satellite_anchor_age_ms`; validation binned by anchor age. The solution must define the max anchor age beyond which estimates degrade to `visual_propagated` / `dead_reckoned` with monotonically growing covariance.
+- **AC-1.4** — Each estimate reports: 95% covariance ellipse semi-major axis (m) AND a label `{satellite_anchored, visual_propagated, dead_reckoned}`.
 
 ## Image Processing Quality
-
-- **AC-2.1** — Image registration rate **>95%** for normal flight segments (defined as: nadir flight ±10° bank / pitch, ≥40% overlap with prior frame, daytime, season-matched satellite tile).
-- **AC-2.2** — Mean Reprojection Error (MRE):
-  - **<1.0 px** for VO frame-to-frame homography on overlapping aerial pairs;
-  - **<2.5 px** for satellite-anchored cross-domain (UAV photo ↔ ortho satellite tile) registration.
+- **AC-2.1a — Frame-to-frame registration**: succeeds for **>95%** of normal flight segments (defined: nadir ±10° bank/pitch, ≥40% prior-frame overlap, daytime, usable texture, no full visual blackout).
+- **AC-2.1b — Satellite-anchor registration**: measured separately from AC-2.1a; must satisfy AC-1.1/1.2 accuracy, AC-2.2 cross-domain MRE, AC-8.2 freshness, AC-8.6 retrieval behaviour.
+- **AC-2.2** — Mean Reprojection Error: **<1.0 px** frame-to-frame; **<2.5 px** satellite-anchored cross-domain.
 
 ## Resilience & Edge Cases
-
-- **AC-3.1** — The system shall correctly continue work in the presence of up to **350 m** outliers between two consecutive photos (caused by airframe tilt up to ±20°).
-- **AC-3.2** — The system shall correctly continue work during sharp turns where the next photo overlaps **<5%** with the previous, drifts **<200 m**, and changes heading **<70°**. Sharp-turn frames are expected to fail VO and shall be handled by satellite-based re-localization (place recognition over the satellite tile cache).
-- **AC-3.3** — The system shall handle **≥3 disconnected segments** per flight, connecting each new segment to the previous trajectory via global descriptor retrieval + RANSAC pose-graph relocalization. This is a core capability, not a degraded mode.
-- **AC-3.4** — When the system cannot determine position for **≥3 consecutive frames AND ≥2 s**, it shall send a re-localization request to the ground station via telemetry. While waiting, it continues VO/IMU dead reckoning and the flight controller uses last known position + IMU extrapolation.
+- **AC-3.1** — Tolerate up to **350 m** outliers between two consecutive photos (airframe tilt up to ±20°).
+- **AC-3.2** — Tolerate sharp turns: <5% overlap, <200 m drift, <70° heading change. Sharp-turn frames may fail frame-to-frame registration; recovery via satellite-reference re-localization.
+- **AC-3.3** — Handle **≥3 disconnected segments** per flight via satellite-reference re-localization. Core capability, not degraded mode.
+- **AC-3.4** — On ≥3 consecutive frames AND ≥2 s without a position, request operator re-loc via telemetry; continue dead-reckoned propagation; FC uses last known + IMU extrapolation.
+- **AC-3.5 — Visual blackout + spoofed GPS** (clouds/occlusion/whiteout while FC reports GPS denial/spoof):
+  - Switch label to `{dead_reckoned}` within ≤1 processed frame OR ≤400 ms.
+  - Reject spoofed GPS as estimator input.
+  - Propagate from last trusted state + FC IMU/attitude/airspeed/altitude until visual or satellite anchoring recovers.
+  - Covariance grows monotonically.
+  - `horiz_accuracy` field of the GPS message to the FC must not under-report the 95% covariance semi-major axis.
+  - `VISUAL_BLACKOUT_IMU_ONLY` STATUSTEXT to QGroundControl at 1–2 Hz.
 
 ## Real-Time Onboard Performance
-
-- **AC-4.1** — End-to-end latency from camera capture to GPS coordinate output to the flight controller shall be **<400 ms p95**. Up to ~10% of frames may be dropped under sustained load (skip-allowed).
-- **AC-4.2** — Memory usage shall remain below **8 GB** shared on Jetson Orin Nano Super (CPU and GPU share the same 8 GB LPDDR5 pool).
-- **AC-4.3** — The system shall output its position estimate to the flight controller via **two parallel MAVLink channels**, both emitted by **pymavlink** (general telemetry uses MAVSDK):
-  - **Primary**: `GPS_INPUT` targeting **ArduPilot** with `GPS1_TYPE=14` (MAVLink GPS substitute). Matches the "replacement for the GPS module" framing of the build.
-  - **Auxiliary** (when the EKF emits a fix with full 6-DoF covariance and quality > VISO_QUAL_MIN): `ODOMETRY` so EKF3 can fuse the richer covariance + native yaw error + quality field. ArduPilot's own dev docs designate ODOMETRY as the preferred external-nav channel for non-GPS substitution; we hybridise to keep AC-4.3's GPS-substitute framing while not throwing away the covariance fidelity that AC-NEW-4 depends on.
-  - FC source priorities are configured so GPS_INPUT remains the failover path if ODOMETRY trips a parameter gate.
-  - **v1 scope clause (added 2026-04-26 — see solution_draft03 finding M-30)**: v1 ships **GPS_INPUT only**; the ODOMETRY auxiliary channel is intentionally **disabled** in v1 because feeding both `GPS_INPUT` and `ODOMETRY` for overlapping axes triggers ArduPilot EKF3 double-fusion bugs (issues #30076 / #32506). `EK3_SRC1_*=GPS+Compass`; ODOMETRY emission re-enables in v1.1 once F-T9 SITL confirms PR #30080-class clean source-switching. Tests therefore assert v1 emits GPS_INPUT only and that ODOMETRY is *intentionally absent* on the wire.
-  - (Decision rationale: MAVSDK has no native GPS_INPUT support — see `_docs/00_research/00_ac_assessment.md` Q-1; ODOMETRY hybrid rationale — see Mode B finding M-1 in `_docs/00_research/02_fact_cards.md`; v1 single-channel rationale — see Mode B round-2 finding M-30 in `_docs/00_research/02_fact_cards.md` / solution_draft03.)
-- **AC-4.4** — Position estimates are streamed to the flight controller frame-by-frame; the system shall not batch or delay output.
-- **AC-4.5** — The system may refine previously calculated positions and send corrections to the flight controller as updated estimates.
+- **AC-4.1** — End-to-end latency (camera capture → GPS to FC) **<400 ms p95**. Up to ~10% frames may drop under sustained load.
+- **AC-4.2** — Memory **<8 GB shared** on Jetson Orin Nano Super.
+- **AC-4.3 — FC output contract**: WGS84 coordinates delivered to each supported FC via that FC's documented external-positioning interface — MAVLink `GPS_INPUT` for ArduPilot Plane, MSP2 `MSP2_SENSOR_GPS` for iNav. Honest covariance is carried in the field each FC uses for outlier rejection (under-reported covariance is a defect, see AC-NEW-4). Source-label semantics per AC-1.4 are emitted out-of-band via the FC-appropriate channel (e.g. MAVLink `STATUSTEXT` / `NAMED_VALUE_FLOAT` for ArduPilot; MSP equivalent for iNav). Where the FC supports it, implementation may also emit an optional auxiliary external-odometry message when the estimator delivers full 6-DoF covariance + quality above a configured threshold. Per-FC parameter wiring (EKF source-set selection on ArduPilot; GPS provider / UART role on iNav), FDR-side message variants, and out-of-band channel choice remain design decisions.
+- **AC-4.4** — Estimates streamed frame-by-frame; no batching/delay.
+- **AC-4.5** — System may refine prior estimates and emit corrections.
 
 ## Startup & Failsafe
-
-- **AC-5.1** — The system shall initialise using the last known valid GPS position from the flight controller's EKF, plus IMU-extrapolated position at the moment of GPS denial.
-- **AC-5.2** — If the system fails to produce any position estimate for **>3 s**, the flight controller shall fall back to IMU-only dead reckoning and the system shall log the failure.
-- **AC-5.3** — On companion computer reboot mid-flight, the system shall attempt to re-initialise from the flight controller's current IMU-extrapolated position. See AC-NEW-1 for the cold-start time-to-first-fix budget.
+- **AC-5.1** — Initialise from FC EKF's last valid GPS + IMU-extrapolated position at GPS denial.
+- **AC-5.2** — On >3 s without estimate, FC falls back to IMU-only dead reckoning; system logs failure. Verify in production param sets of each supported FC (ArduPilot Plane SITL + iNav SITL or equivalent).
+- **AC-5.3** — On companion reboot mid-flight, re-initialise from FC's current IMU-extrapolated position. Cold-start TTFF in AC-NEW-1.
 
 ## Ground Station & Telemetry
-
-- **AC-6.1** — Position estimates and confidence scores shall be streamed to **QGroundControl** via the MAVLink telemetry link. High-rate (per-frame) content stays on the local link for forensics; the GCS link is downsampled to **1–2 Hz** for situational awareness.
-- **AC-6.2** — The ground station can send commands to the onboard system (e.g., operator-assisted re-localization hint with approximate coordinates) via STATUSTEXT, NAMED_VALUE_FLOAT, or a custom MAVLink dialect.
-- **AC-6.3** — Output coordinates are in **WGS84** format (matches GPS_INPUT spec).
+- **AC-6.1** — Position estimates + confidence stream to QGroundControl over MAVLink at **1–2 Hz** downsampled (high-rate stays on local FDR).
+- **AC-6.2** — GCS may send commands (e.g., operator re-loc hint) via standard MAVLink (`STATUSTEXT`, `NAMED_VALUE_FLOAT`) or a custom dialect.
+- **AC-6.3** — Output coordinates in WGS84.
 
 ## Object Localization (AI Camera)
-
-- **AC-7.1** — Other onboard AI systems may request GPS coordinates of objects detected by the AI camera. Localization accuracy is **consistent with the frame-center accuracy of the GPS-Denied system in level flight (bank/pitch <5°)**. In maneuvering flight, ground-projection error is bounded by `altitude × |sin(unknown_bank_or_pitch)|` and the system shall publish that bound alongside the estimate.
-- **AC-7.2** — The system computes object coordinates trigonometrically using: current UAV GPS position (from GPS-Denied), known AI-camera gimbal angle, zoom, and current flight altitude. Flat-terrain assumption applies.
+- **AC-7.1** — AI systems may request GPS for AI-camera-detected objects. Accuracy consistent with frame-center accuracy in level flight (bank/pitch <5°). In maneuvering flight, error bounded by `altitude × |sin(unknown_bank_or_pitch)|` and that bound is published alongside the estimate.
+- **AC-7.2** — Object coordinates computed trigonometrically from current UAV position, AI-camera gimbal angle, zoom, and altitude. Flat-terrain assumption.
 
 ## Satellite Reference Imagery
+- **AC-8.1** — Imagery via Azaion Suite Satellite Service (offline cache interface; no direct commercial-provider calls). Cache-interface resolution ≥0.5 m/px, ideally 0.3 m/px.
+- **AC-8.2** — Tile freshness: <6 mo (active-conflict sectors), <12 mo (stable rear). Older → reject or downgrade (AC-NEW-6).
+- **AC-8.3** — Imagery pre-loaded onto companion before flight; offline preprocessing time not time-critical. Pre-extracted descriptors/indices count against the cache budget unless explicitly carved out.
+- **AC-8.4** — Mid-flight tile generation: continuously orthorectify nav-camera frames into basemap-projected tiles, deduplicated (latest/highest-quality wins). Tiles are written **only** to the local cache while airborne — in-air outbound writes to `satellite-provider` are **forbidden** for drone-security reasons; enforced by a `flight state` process-level gate (see `architecture.md`). Upload to `satellite-provider` happens **only after landing**, triggered by a separate operator-side post-landing upload tool. Local mid-flight tile format matches `satellite-provider`'s on-disk format so post-landing upload is byte-identical. Each uploaded tile carries quality metadata sufficient for the Service's ingest pipeline (AC-NEW-7).
+- **AC-8.5** — No raw nav-camera or AI-camera frames retained in normal operation; tiles are the only persistent imagery. Forensic exception: ≤0.1 Hz thumbnail log of frames that failed tile generation, within FDR budget (AC-NEW-3).
+- **AC-8.6 — Satellite-anchor relocalization robustness**:
+  - **Scale-ratio**: any UAV-frame ground footprint at the deployment altitude band must be retrievable from the cache regardless of internal tiling/indexing.
+  - **Scene change in active-conflict sectors**: cratering / building destruction / road realignment must not collapse retrieval recall, measured against a labelled change-pair dataset over season-matched tiles. No `satellite_anchored` label on stale-tile match (per AC-NEW-6).
+  - **Compute & latency**: relocalization must remain inside AC-4.1 latency + AC-4.2 memory budgets under both steady-state and re-loc-trigger workloads.
 
-- **AC-8.1** — Satellite reference imagery is provided by the **Azaion Suite Satellite Service** (a separate component of the Suite). The runtime onboard system consumes this service through an offline tile cache interface; it does **not** call commercial providers (Maxar, Airbus, Planet, etc.) directly. The Satellite Service is responsible for upstream sourcing and is out of scope for this build. Required resolution at the cache interface: **at least 0.5 m/pixel, ideally 0.3 m/pixel**.
-- **AC-8.2** — Satellite tiles consumed at runtime shall be:
-  - **<6 months old** for active-conflict sectors;
-  - **<12 months old** for stable rear sectors.
-  System shall reject or downgrade-confidence on tiles older than these thresholds (see AC-NEW-6).
-- **AC-8.3** — Satellite imagery for the operational area shall be **pre-loaded and pre-processed** onto the companion computer before flight. Offline preprocessing time is not time-critical (minutes/hours). Pre-extracted tile descriptors (e.g., SuperPoint keypoints/descriptors and DINOv2-VLAD global descriptors) are part of the cache.
-- **AC-8.4** — **Mid-flight tile generation & write-back**: during flight, the system shall continuously orthorectify navigation-camera frames into tiles aligned with the basemap projection and store them in the local cache, **deduplicated** so each ground sector is stored at most once (latest / highest-quality tile wins). On landing, the companion computer shall upload newly generated tiles back to the Azaion Suite Satellite Service so that the next mission cache contains imagery refreshed by the previous flight.
-- **AC-8.5** — **Storage policy**: the system shall **not** retain raw navigation-camera frames or AI-camera frames as part of normal operation. Tiles are the only persistent imagery artifact. Forensic exception: a low-rate (≤0.1 Hz) thumbnail log of frames that **failed** tile generation may be retained for debugging within the FDR budget (AC-NEW-3).
-- **AC-8.6** — **VPR retrieval unit + change-robustness**:
-  - The Visual Place Recognition (Component 2) FAISS index shall be built over **ground-footprint-sized "VPR chunks"** (~600–800 m at the deployment altitude band, with **40–50 % overlap** between adjacent chunks), **decoupled from the slippy-XYZ storage tile** (z=20). Any UAV frame footprint shall fall fully inside ≥1 chunk regardless of position.
-  - The index shall be **multi-scale**: in addition to fine-scale chunks (derived from z=20 storage), a coarser-scale chunk descriptor set (z=17 or z=18 effective scale) shall be maintained for change-robust retrieval in **active-conflict sectors** where building destruction or major scene change is expected.
-  - VPR top-K shall be **dynamically sized** by sector classification (AC-NEW-6) and EKF position covariance: K=5 in stable sectors with σ_xy ≤ 20 m; K=20 in active-conflict sectors; K=50 on expanding-window fallback.
-  - VPR shall be **invoked conditionally**, not on every frame: in steady state (last anchor age < 2 s, σ_xy < 20 m, VO healthy), the system uses a geometric prior from IMU+VO predicted position to rank candidate chunks by distance alone. VPR's DINOv2 forward is invoked on **re-loc triggers** (cold start AC-NEW-1, sharp turn AC-3.2, disconnected segment AC-3.3, σ_xy > 50 m, or VO failure for ≥2 frames).
+## Additional AC
 
-## New AC (added in Phase 1 assessment, expanded with rationale & validation)
-
-### AC-NEW-1 — Time-to-first-fix on cold start
-
-**Statement.** From companion-computer boot, the system shall emit its first valid `GPS_INPUT` message in **<30 s**, given an IMU-extrapolated initial position handed over from the flight controller's EKF.
-
-**Why it matters.** A mid-flight reboot (brown-out, watchdog reset, OS panic) is a realistic scenario on a fixed-wing UAV running an 8-hour mission. The autopilot continues to fly on IMU dead reckoning during the gap; a 30 s budget keeps that drift under ~500 m at 60 km/h cruise, which the EKF can absorb when our first fix arrives.
-
-**Implementation drivers.** TensorRT engines must be built at install time (not at first run); CUDA / TRT init <5 s; tile-cache mmap warm at start; FAISS index loaded before MAVLink connect; first VPR retrieval + cross-view match must succeed at full resolution within the remaining budget.
-
-**Validation.** Bench: cold-boot the companion 50× with simulated FC-pose input; record time from boot to first valid `GPS_INPUT` MAVLink frame. Pass = 95% percentile <30 s.
+### AC-NEW-1 — Cold-start TTFF
+**Statement.** From companion boot, first valid external-position MAVLink frame **<30 s p95**, given an IMU-extrapolated initial position from FC EKF.
+**Why.** Mid-flight reboot is realistic on 8 h missions; FC dead-reckons during the gap, ~500 m drift max at 60 km/h.
+**Validation.** Cold-boot 50× with simulated FC pose; measure boot → first frame; pass = 95th percentile <30 s.
 
 ### AC-NEW-2 — Spoofing-promotion latency
-
-**Statement.** When the flight controller signals GPS denial or spoofing (ArduPilot fix-loss / EKF lane-switch event; PX4 `EKF2_GPS_SPOOFED` flag if PX4 ever returns to scope), the GPS-Denied system shall promote its own estimate to the FC's primary GPS source within **<3 s**.
-
-**Why it matters.** Without this gate, the FC may continue to follow a spoofed real-GPS source while our valid estimate sits idle. 3 s is short enough to keep the FC from acting on a malicious heading change but long enough to ride out a single-frame anomaly.
-
-**Implementation drivers.** Subscribe to `GPS_RAW_INT`, `EKF_STATUS_REPORT`, `SYS_STATUS`. Maintain an internal "real-GPS health" rolling average; switch to "primary" mode (raise our `GPS_INPUT` `fix_type` to 3D and assert) when health drops below threshold for ≥1 s. Emit `STATUSTEXT` to QGC on every promotion / demotion.
-
-**Validation.** SITL: simulate spoofing (inject false `GPS_RAW_INT` from a malicious node); measure time from spoof onset to our promotion. Pass = 95% percentile <3 s.
+**Statement.** When FC signals GPS denial/spoof, promote onboard estimate to FC's primary position source within **<3 s p95**.
+**Why.** Without this, FC may follow a spoofed source while a valid onboard estimate sits idle; 3 s rides out one-frame anomalies but blocks malicious heading changes.
+**Validation.** SITL on each supported FC (ArduPilot Plane + iNav, production param sets): inject false GPS, measure spoof onset → promotion; pass = 95th percentile <3 s on both.
 
 ### AC-NEW-3 — Flight Data Recorder
-
-**Statement.** The system shall retain to non-volatile storage, per flight: per-frame position estimates with covariance and source-label, IMU traces from the FC at full rate, all emitted `GPS_INPUT` frames, MAVLink raw stream (tlog), system health (CPU / GPU / temp / throttle), tiles generated mid-flight (AC-8.4), and a low-rate (≤0.1 Hz) thumbnail log of frames that failed tile generation. **Raw nav-cam frames and AI-cam frames are NOT retained** (AC-8.5). Storage cap **64 GB / flight**; recorder rolls over (oldest segment dropped first) after cap.
-
-**Why it matters.** Tiles, telemetry traces, and IMU are the operationally useful artifacts: they reproduce the mission, feed the next mission's cache (AC-8.4), and let post-mission analysis explain any false-position event (AC-NEW-4). Raw frames are large and redundant once tiles exist.
-
-**Implementation drivers.** Per-day directory layout; fixed-size segment files; rollover policy on segment-close, not on every write. NVMe ≥64 GB on top of the persistent satellite-tile cache.
-
-**Validation.** Bench: run an 8-hour synthetic load (3 Hz nav frames replayed from disk), assert the FDR ends ≤64 GB and no payload class is silently dropped without a logged rollover event.
+**Statement.** Per flight, retain to NVM: per-frame estimates with covariance + source-label; FC IMU traces (full rate); all emitted external-position MAVLink frames; raw MAVLink stream (tlog); system health (CPU/GPU/temp/throttle); mid-flight tiles (AC-8.4); ≤0.1 Hz thumbnail log of failed tile-gen frames. **No raw nav-cam/AI-cam frames** (AC-8.5). Cap **64 GB / flight**; oldest segment dropped first on rollover.
+**Why.** Tiles + telemetry + IMU reproduce the mission, feed next mission's cache (AC-8.4), explain false-position events (AC-NEW-4). Raw frames are large + redundant once tiles exist.
+**Validation.** 8 h synthetic load (3 Hz nav frames replayed); assert FDR ≤64 GB; no payload class silently dropped without a logged rollover.
 
 ### AC-NEW-4 — False-position safety budget
-
-**Statement.**
-- P(reported estimate error > **500 m**) **<0.1 %** per flight.
-- P(reported estimate error > **1 km**) **<0.01 %** per flight.
-
-**Why it matters.** A single 1-km-off `GPS_INPUT` frame can hand the FC a heading that flies the UAV outside the geofence in seconds. The covariance carried in `GPS_INPUT` (`h_acc`) is the FC's only defense; this AC bounds the **probability** of our covariance under-reporting reality.
-
-**Implementation drivers.** EKF covariance must be calibrated, not optimistic. Cross-view fixes with low inlier ratio must be **rejected**, not down-weighted to "small but non-zero". Outlier rejection at the EKF stage (Mahalanobis gate) is mandatory.
-
-**Validation.** Monte Carlo over the AerialVL public dataset (S03) and our own recorded Mavic flights, with synthetic IMU injection where applicable; report error CDF; pass = both probabilities below budget across ≥100 simulated flights worth of frames.
+**Statement.** Per flight: **P(error >500 m) <0.1 %**, **P(error >1 km) <0.01 %**.
+**Why.** A single 1-km-off frame can fly the UAV outside the geofence; covariance carried in the MAVLink message is the FC's only defense.
+**Validation.** Monte Carlo over the currently-available data corpus (Derkachi flight + 60 stills + synthetic perturbations); report error CDF with stated 95% confidence interval; pass = both probabilities below budget within the CI's lower bound. Multi-flight statistical headroom (originally framed as ≥100 flights) is residual risk tracked in the Step 4 risk register; **D-PROJ-3** reopens this validation when additional multi-flight data becomes available.
 
 ### AC-NEW-5 — Operational environmental envelope
-
-**Statement.** Operating temperature **−20 °C to +50 °C**; vibration / shock per RTCA DO-160G low-altitude UAV-class envelope. The cooling solution shall sustain the **25 W** power mode at the upper temperature bound for the full **8-hour duty cycle** without thermal throttling.
-
-**Why it matters.** Without this, all latency / accuracy ACs are conditional on a benign thermal day. Eastern/southern Ukraine summers easily exceed +35 °C ambient inside a UAV bay; without active cooling, the Jetson throttles to 15 W mode and our 400 ms latency budget collapses.
-
-**Implementation drivers.** Forced-air or active heatsink sized for 25 W continuous at +50 °C ambient bay temperature; thermal sensors logged in FDR (AC-NEW-3); throttle event = automatic `STATUSTEXT` warning to QGC.
-
-**Validation.** Hot-soak chamber test: 25 W workload at +50 °C ambient for 8 h; assert no throttle. Cold-soak: −20 °C cold-start to first fix within AC-NEW-1 budget.
+**Statement.** Operating temp **−20 °C to +50 °C**; vibration/shock per RTCA DO-160G low-altitude UAV-class. Cooling sustains **25 W** at the upper temp for the full **8-hour duty cycle** without throttling.
+**Why.** Without this, all latency/accuracy AC are conditional on a benign thermal day; +35 °C bay temps cause Jetson to throttle to 15 W, collapsing the 400 ms latency budget.
+**Validation.** Hot-soak: 25 W @ +50 °C for 8 h, no throttle. Cold-soak: −20 °C cold-start within AC-NEW-1.
 
 ### AC-NEW-6 — Imagery freshness enforcement
-
-**Statement.** The system shall reject (or downgrade confidence on) any satellite tile whose capture date violates AC-8.2 (>6 months old in active-conflict sectors; >12 months old in stable rear sectors). Tiles generated mid-flight (AC-8.4) and not yet uploaded to the Suite Satellite Service are timestamped with the current flight date and treated as fresh.
-
-**Why it matters.** Stale satellite tiles are the dominant cross-view-matching failure mode in active-conflict sectors (cratering, dam destruction, road realignment). A confident match against a stale tile is worse than no match.
-
-**Implementation drivers.** Each tile carries `capture_date` metadata in the cache index. Sector classification (active vs stable) is part of the operational area definition handed in pre-flight. Confidence weight = 1.0 if within freshness budget, linearly decayed to 0.0 over a 30-day grace zone past the budget, hard reject beyond the grace.
-
-**Validation.** Inject tiles with synthetic age into the cache; verify rejection / decay curve matches spec; verify a stale-tile match never produces a `satellite_anchored` source label.
+**Statement.** System rejects (or downgrades) any tile whose capture date violates AC-8.2. Mid-flight tiles (AC-8.4) not yet uploaded are timestamped current and treated as fresh.
+**Why.** Stale tiles are the dominant cross-view-matching failure mode in active-conflict sectors; a confident match on a stale tile is worse than no match.
+**Validation.** Inject synthetic-age tiles; verify rejection/decay matches spec; verify stale-tile match never produces `satellite_anchored`.
 
 ### AC-NEW-7 — Cache-poisoning safety budget
+**Statement.** Per flight, across all onboard tiles written (AC-8.4): **P(geo-misalign >30 m) <1 %**, **P(>100 m) <0.1 %**.
+**Why.** Onboard tiles feed back into the `satellite-provider` basemap when uploaded post-landing (AC-8.4). A bad onboard pose with optimistic covariance writes a misaligned tile that becomes the next flight's anchor — cross-flight error compounding that AC-NEW-4 doesn't capture.
+**External-dependency note.** The parent-suite `satellite-provider` is expected to operate a multi-flight ingest-side trust/voting layer that gates onboard-tile promotion to "trusted basemap" until multiple independent flights agree on geo-alignment. The ingest endpoint and voting layer are **not currently implemented in `satellite-provider`** and are tracked as a parent-suite design task (**D-PROJ-2**). Onboard's job (AC-8.4) is to publish per-tile quality metadata sufficient for that layer. End-to-end AC-NEW-7 evidence depends on the `satellite-provider` contract being added.
+**Validation.** Onboard-only Monte Carlo replay over the currently-available data corpus + synthetic over-confidence injection (deflate covariance ×1.5–3); report error CDF with stated 95% confidence interval; pass = both probabilities below budget within the CI's lower bound for the onboard-side contribution. Multi-flight statistical headroom and the `satellite-provider` voting-side contract verification are residual risks tracked in the Step 4 risk register; **D-PROJ-3** reopens onboard validation when additional multi-flight data becomes available; **D-PROJ-2** reopens cross-suite validation once the ingest + voting layer is built.
 
-**Statement.** Per flight, across all onboard tiles written by Component 1b (in-flight ortho-tile generator):
-
-- P(onboard tile geo-misaligned > **30 m**) **<1 %**.
-- P(onboard tile geo-misaligned > **100 m**) **<0.1 %**.
-
-**Why it matters.** Onboard tiles feed back into the Suite Satellite Service's basemap (AC-8.4). Without this AC, a confidently-bad EKF pose can write a misaligned tile that, after Service ingest, becomes the next flight's satellite anchor — producing cross-flight error compounding that AC-NEW-4 (single-flight false-position budget) does not capture. This AC bounds the **probability** that an onboard tile's claimed geo-alignment is wrong by a margin that would propagate to a downstream flight.
-
-**Implementation drivers.**
-- Service-source tiles are immutable within freshness budget (AC-8.2); onboard tiles overwrite only stale or other-onboard tiles.
-- The Suite Satellite Service ingest applies a **2-flight voting layer**: an onboard tile gets promoted to "trusted basemap" only after **N≥2 independent flights** confirm consistent geo-alignment within X m of each other. (Active sectors per AC-NEW-6 may use single-flight promotion when σ_xy ≤ 3 m AND OSM-road-overlap ≥ 70 %.)
-- The Component-1b parent-pose covariance is a **hard gate** in the local quality score: σ_xy ≤ 5 m for a hard write (`trust_level = candidate`); σ_xy ≤ 3 m for `trust_level = candidate` with full quality; tiles written in the σ_xy ∈ (3, 5] m band are marked `trust_level = soft` in the sidecar.
-- Eligibility check (Component 1b) tightens generation gate from σ_xy ≤ 10 m to σ_xy ≤ 5 m.
-
-**Validation.** Multi-flight Monte Carlo replay over AerialVL + Mavic + AerialExtreMatch with **synthetic over-confidence injection** (artificially deflate EKF covariance by 1.5×–3×): assert both probabilities below budget across ≥100 simulated flights worth of frames. Independently, Service-side voting layer is exercised in F-T3 to verify candidate tiles are not promoted to trusted basemap before N-flight confirmation.
+### AC-NEW-8 — Visual blackout + GPS spoofing degraded mode
+**Statement.** When the navigation camera is fully unusable AND FC reports GPS denial/spoof:
+- continue emitting external-position MAVLink frames from IMU-only propagation for **≤30 s** after the last trusted anchor (or until covariance trips fail threshold);
+- label every estimate `{dead_reckoned}`; degrade MAVLink fix-quality to "2D fix or worse" when 95% covariance semi-major axis **>100 m**;
+- escalate to "no fix" (`horiz_accuracy=999.0`) + `VISUAL_BLACKOUT_FAILSAFE` STATUSTEXT when 95% covariance >**500 m** OR blackout >**30 s** without a trusted re-anchor;
+- never promote spoofed real-GPS back into the estimator unless FC GPS health stable + non-spoofed for **≥10 s** AND a visual/satellite consistency check has succeeded.
+**Why.** During cloud/whiteout + spoofing, no honest correction is available; only safe behaviour is IMU-only dead reckoning with rapidly-growing uncertainty, never pretending stale visual or spoofed GPS remains valid.
+**Validation.** SITL/replay on each FC: inject 5 s / 15 s / 35 s blackouts while spoofing GPS; assert mode transition ≤400 ms, spoofed GPS ignored, covariance grows monotonically, MAVLink fields degrade at thresholds, recovery only via trusted anchor or 10-s GPS-health + visual-consistency gate.

_docs/00_problem/input_data/data_parameters.md

@@ -1,8 +1,2 @@
-- Height
-  - 400m
-- Camera:
-  - Name: ADTi Surveyor Lite 26S v2
-  - Resolution: 26MP
-  - Image resolution: 6252*4168
-  - Focal length: 25mm
-  - Sensor width: 23.5
+- Height: 400m
+- Camera: ADTi Surveyor Lite 20MP 20L V1

_docs/00_problem/input_data/expected_results/results_report.md

@@ -1,166 +1,97 @@
-# Expected Results
+# Expected Results Mapping
 
-Maps every input data item to its quantifiable expected result.
-Tests use this mapping to compare actual system output against known-correct answers.
+## Scope
 
-## Result Format Legend
+`coordinates.csv` is the current source of truth for the provided still-image nadir set. It gives expected WGS84 frame-center coordinates for `AD000001.jpg` through `AD000060.jpg`.
 
-| Result Type | When to Use | Example |
-|-------------|-------------|---------|
-| Exact value | Output must match precisely | `fix_type: 3`, `satellites_visible: 10` |
-| Tolerance range | Numeric output with acceptable variance | `lat: 48.275292 ± 50m` |
-| Threshold | Output must exceed or stay below a limit | `latency < 400ms`, `memory < 8GB` |
-| Pattern match | Output must match a string/regex pattern | `RELOC_REQ: last_lat=.* last_lon=.* uncertainty=.*m` |
-| File reference | Complex output compared against a reference file | `match expected_results/position_accuracy.csv` |
-| Set/count | Output must contain specific items or counts | `registered_frames / total_frames > 0.95` |
+This data is sufficient for black-box frame-center geolocation tests against still images. The Derkachi representative fixture in `input_data/flight_derkachi/` adds cropped nadir video plus synchronized `SCALED_IMU2` and `GLOBAL_POSITION_INT` telemetry. It is sufficient for fixture validation, video/telemetry synchronization, replay, latency, VIO smoke tests, and trajectory comparison against the tlog GPS path. It is not sufficient by itself for final production accuracy because raw camera calibration, lens distortion, and exact camera-to-body calibration are still pending.
 
-## Comparison Methods
+## Pass / Fail Rules
 
-| Method | Description | Tolerance Syntax |
-|--------|-------------|-----------------|
-| `numeric_tolerance` | abs(actual - expected) ≤ tolerance | `± <value>` |
-| `threshold_min` | actual ≥ threshold | `≥ <value>` |
-| `threshold_max` | actual ≤ threshold | `≤ <value>` |
-| `percentage` | percentage of items meeting criterion | `≥ N%` |
-| `exact` | actual == expected | N/A |
-| `regex` | actual matches regex pattern | regex string |
-| `file_reference` | compare against reference file | file path |
+- **Normal frame-center geolocation**: estimated frame center is within 50 m of the expected WGS84 coordinate.
+- **Stretch accuracy bin**: estimated frame center is within 20 m of the expected WGS84 coordinate.
+- **Dataset aggregate**: at least 80% of mapped images pass the 50 m threshold and at least 50% pass the 20 m threshold.
+- **Output shape**: each result must include image name, estimated `lat`, estimated `lon`, error in meters, source label, 95% covariance semi-major axis, and `last_satellite_anchor_age_ms`.
 
-## Input → Expected Result Mapping
+## Input To Expected Output Map
 
-### Position Accuracy (60-image flight sequence)
+### Still-Image Frame Centers
 
-Ground truth GPS coordinates for each frame are in `coordinates.csv`. The system processes these frames sequentially (simulating a real flight) with corresponding IMU data (200Hz, from SITL ArduPilot or synthetic generation from trajectory) and satellite tile matches. The system outputs estimated GPS coordinates per frame. Expected results compare estimated positions against ground truth.
+| Input image | Expected latitude | Expected longitude | Primary threshold | Stretch threshold |
+|-------------|-------------------|--------------------|-------------------|-------------------|
+| AD000001.jpg | 48.275292 | 37.385220 | <= 50 m | <= 20 m |
+| AD000002.jpg | 48.275001 | 37.382922 | <= 50 m | <= 20 m |
+| AD000003.jpg | 48.274520 | 37.381657 | <= 50 m | <= 20 m |
+| AD000004.jpg | 48.274956 | 37.379004 | <= 50 m | <= 20 m |
+| AD000005.jpg | 48.273997 | 37.379828 | <= 50 m | <= 20 m |
+| AD000006.jpg | 48.272538 | 37.380294 | <= 50 m | <= 20 m |
+| AD000007.jpg | 48.272408 | 37.379153 | <= 50 m | <= 20 m |
+| AD000008.jpg | 48.271992 | 37.377572 | <= 50 m | <= 20 m |
+| AD000009.jpg | 48.271376 | 37.376671 | <= 50 m | <= 20 m |
+| AD000010.jpg | 48.271233 | 37.374806 | <= 50 m | <= 20 m |
+| AD000011.jpg | 48.270334 | 37.374442 | <= 50 m | <= 20 m |
+| AD000012.jpg | 48.269922 | 37.373284 | <= 50 m | <= 20 m |
+| AD000013.jpg | 48.269366 | 37.372134 | <= 50 m | <= 20 m |
+| AD000014.jpg | 48.268759 | 37.370940 | <= 50 m | <= 20 m |
+| AD000015.jpg | 48.268291 | 37.369815 | <= 50 m | <= 20 m |
+| AD000016.jpg | 48.267719 | 37.368469 | <= 50 m | <= 20 m |
+| AD000017.jpg | 48.267461 | 37.367255 | <= 50 m | <= 20 m |
+| AD000018.jpg | 48.266663 | 37.365888 | <= 50 m | <= 20 m |
+| AD000019.jpg | 48.266135 | 37.365460 | <= 50 m | <= 20 m |
+| AD000020.jpg | 48.265574 | 37.364211 | <= 50 m | <= 20 m |
+| AD000021.jpg | 48.264892 | 37.362998 | <= 50 m | <= 20 m |
+| AD000022.jpg | 48.264393 | 37.361086 | <= 50 m | <= 20 m |
+| AD000023.jpg | 48.263803 | 37.361028 | <= 50 m | <= 20 m |
+| AD000024.jpg | 48.263014 | 37.359878 | <= 50 m | <= 20 m |
+| AD000025.jpg | 48.262635 | 37.358277 | <= 50 m | <= 20 m |
+| AD000026.jpg | 48.261819 | 37.357116 | <= 50 m | <= 20 m |
+| AD000027.jpg | 48.261182 | 37.355907 | <= 50 m | <= 20 m |
+| AD000028.jpg | 48.260727 | 37.354723 | <= 50 m | <= 20 m |
+| AD000029.jpg | 48.260117 | 37.353469 | <= 50 m | <= 20 m |
+| AD000030.jpg | 48.259677 | 37.352165 | <= 50 m | <= 20 m |
+| AD000031.jpg | 48.258881 | 37.351376 | <= 50 m | <= 20 m |
+| AD000032.jpg | 48.258425 | 37.349964 | <= 50 m | <= 20 m |
+| AD000033.jpg | 48.258653 | 37.347004 | <= 50 m | <= 20 m |
+| AD000034.jpg | 48.257879 | 37.347711 | <= 50 m | <= 20 m |
+| AD000035.jpg | 48.256777 | 37.348444 | <= 50 m | <= 20 m |
+| AD000036.jpg | 48.255756 | 37.348098 | <= 50 m | <= 20 m |
+| AD000037.jpg | 48.255375 | 37.346549 | <= 50 m | <= 20 m |
+| AD000038.jpg | 48.254799 | 37.345603 | <= 50 m | <= 20 m |
+| AD000039.jpg | 48.254557 | 37.344566 | <= 50 m | <= 20 m |
+| AD000040.jpg | 48.254380 | 37.344375 | <= 50 m | <= 20 m |
+| AD000041.jpg | 48.253722 | 37.343093 | <= 50 m | <= 20 m |
+| AD000042.jpg | 48.254205 | 37.340532 | <= 50 m | <= 20 m |
+| AD000043.jpg | 48.252380 | 37.342112 | <= 50 m | <= 20 m |
+| AD000044.jpg | 48.251489 | 37.343079 | <= 50 m | <= 20 m |
+| AD000045.jpg | 48.251085 | 37.346128 | <= 50 m | <= 20 m |
+| AD000046.jpg | 48.250413 | 37.344034 | <= 50 m | <= 20 m |
+| AD000047.jpg | 48.249414 | 37.343296 | <= 50 m | <= 20 m |
+| AD000048.jpg | 48.249114 | 37.346895 | <= 50 m | <= 20 m |
+| AD000049.jpg | 48.250241 | 37.347741 | <= 50 m | <= 20 m |
+| AD000050.jpg | 48.250974 | 37.348379 | <= 50 m | <= 20 m |
+| AD000051.jpg | 48.251528 | 37.349468 | <= 50 m | <= 20 m |
+| AD000052.jpg | 48.251873 | 37.350485 | <= 50 m | <= 20 m |
+| AD000053.jpg | 48.252161 | 37.351491 | <= 50 m | <= 20 m |
+| AD000054.jpg | 48.252685 | 37.352343 | <= 50 m | <= 20 m |
+| AD000055.jpg | 48.253268 | 37.353119 | <= 50 m | <= 20 m |
+| AD000056.jpg | 48.253767 | 37.354246 | <= 50 m | <= 20 m |
+| AD000057.jpg | 48.254329 | 37.354946 | <= 50 m | <= 20 m |
+| AD000058.jpg | 48.254874 | 37.355765 | <= 50 m | <= 20 m |
+| AD000059.jpg | 48.255481 | 37.356501 | <= 50 m | <= 20 m |
+| AD000060.jpg | 48.256246 | 37.357485 | <= 50 m | <= 20 m |
 
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 1 | `coordinates.csv` (all 60 frames) | Sequential flight images with ground truth GPS | ≥ 80% of frames have position error < 50m from ground truth | percentage | ≥ 80% of frames within 50m | `expected_results/position_accuracy.csv` |
-| 2 | `coordinates.csv` (all 60 frames) | Sequential flight images with ground truth GPS | ≥ 50% of frames have position error < 20m from ground truth (per AC-1.2) | percentage | ≥ 50% of frames within 20m | `expected_results/position_accuracy.csv` |
-| 3 | `coordinates.csv` (all 60 frames) | Sequential flight images with ground truth GPS | Per-frame position output in WGS84 (lat, lon) | numeric_tolerance | each frame ± 100m max (no single frame exceeds 100m error) | `expected_results/position_accuracy.csv` |
-| 4 | `coordinates.csv` (all 60 frames) | Sequential flight images with ground truth GPS | Cumulative VO drift between satellite anchors < 100m | threshold_max | ≤ 100m drift between anchors | N/A |
+### Representative Derkachi Video/IMU Fixture
 
-### GPS_INPUT Message Correctness
+| Input fixture | Expected validation result | Threshold |
+|---------------|----------------------------|-----------|
+| `flight_derkachi/data_imu.csv` | Telemetry CSV has required `timestamp(ms)`, `Time`, `SCALED_IMU2.*`, and `GLOBAL_POSITION_INT.*` columns; non-empty rows are monotonic from `Time=0.0` to `489.9` | 0 missing required columns; 0 decreasing timestamps; 4,900 nonblank rows |
+| `flight_derkachi/flight_derkachi.mp4` | Video stream is readable as cropped nadir footage for replay | H.264, 880 x 720, 30 fps, approximately 490.07 s |
+| Video/telemetry alignment | Video has 14,700 frames and telemetry has 4,900 rows | Exactly 3 video frames per telemetry row; duration delta <=250 ms |
+| Derkachi trajectory comparison | Replay output can be compared to `GLOBAL_POSITION_INT.lat`, `GLOBAL_POSITION_INT.lon`, `GLOBAL_POSITION_INT.alt`, `GLOBAL_POSITION_INT.relative_alt`, velocity, and heading | Thresholds are calibration-gated; use for smoke/relative trajectory validation until intrinsics and camera-to-body calibration are pinned |
 
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 5 | Single frame + IMU data | Normal tracking frame with recent satellite match | `fix_type: 3`, `horiz_accuracy: 5-20m`, `satellites_visible: 10`, lat/lon populated | exact (fix_type, sat), numeric_tolerance (accuracy) | fix_type == 3, horiz_accuracy ∈ [1, 50] | N/A |
-| 6 | Frame sequence, no satellite match for >30s | VO-only tracking, no recent satellite anchor | `fix_type: 3`, `horiz_accuracy: 20-50m` | exact (fix_type), range (accuracy) | fix_type == 3, horiz_accuracy ∈ [20, 100] | N/A |
-| 7 | Frame sequence, VO lost + no satellite | IMU-only dead reckoning | `fix_type: 2`, `horiz_accuracy: 50-200m+` (growing over time) | exact (fix_type), threshold_min (accuracy) | fix_type == 2, horiz_accuracy ≥ 50 | N/A |
-| 8 | VO lost + 3 consecutive satellite failures | Total position failure | `fix_type: 0`, `horiz_accuracy: 999.0` | exact | fix_type == 0, horiz_accuracy == 999.0 | N/A |
-| 9 | Any valid frame | GPS_INPUT output rate | GPS_INPUT messages at 5-10Hz continuous | range | 5 ≤ rate_hz ≤ 10 | N/A |
+## Known Gaps
 
-### Confidence Tier Transitions
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 10 | Frame with satellite match <30s ago, covariance <400m² | HIGH confidence conditions | Confidence tier: HIGH, SSE confidence: "HIGH" | exact | N/A | N/A |
-| 11 | Frame with cuVSLAM OK, no satellite match >30s | MEDIUM confidence conditions | Confidence tier: MEDIUM, SSE confidence: "MEDIUM" | exact | N/A | N/A |
-| 12 | Frame with cuVSLAM lost, IMU-only | LOW confidence conditions | Confidence tier: LOW, SSE confidence: "LOW" | exact | N/A | N/A |
-| 13 | 3+ consecutive total failures | FAILED conditions | Confidence tier: FAILED, SSE confidence: "FAILED", fix_type: 0 | exact | N/A | N/A |
-
-### Image Registration & Visual Odometry
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 14 | 60 sequential flight images | Normal flight (no sharp turns) | Image registration rate ≥ 95% (≥ 57 of 60 registered) | percentage | ≥ 95% | N/A |
-| 15 | 60 sequential flight images | Normal flight images | Mean reprojection error < 1.0 pixels | threshold_max | MRE < 1.0 px | N/A |
-
-### Disconnected Route Segments & Sharp Turns
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 16 | Frames 32-43 from coordinates.csv | Trajectory with direction change (turn area) | System continues producing position estimates through the turn | threshold_min | ≥ 1 position output per frame | N/A |
-| 17 | Simulated consecutive frames with 350m gap | Outlier between 2 consecutive photos due to tilt | System handles outlier, position estimate not corrupted (error < 100m for next valid frame) | threshold_max | ≤ 100m error after recovery | N/A |
-| 18 | Simulated sharp turn (no overlap, <5% overlap, <70° angle, <200m drift) | Sharp turn where VO fails | Satellite re-localization triggers, position recovered within 3 frames after turn | threshold_max | position error ≤ 50m after re-localization | N/A |
-| 19 | Simulated VO loss + satellite match success | Tracking loss → re-localization | cuVSLAM restarts, Component 5 calibrator emits a satellite-anchored fix, FC EKF3 reconverges, tracking_state returns to NORMAL | exact | tracking_state == NORMAL after recovery | N/A |
-
-### 3-Consecutive-Failure Re-Localization
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 20 | Simulated VO loss + 3 satellite match failures | Cannot determine position by any means | Re-localization request sent: `RELOC_REQ: last_lat=.* last_lon=.* uncertainty=.*m` | regex | message matches pattern | N/A |
-| 21 | Re-localization request active | System waiting for operator | GPS_INPUT fix_type=0, system continues IMU prediction, continues satellite matching attempts | exact (fix_type) | fix_type == 0 | N/A |
-| 22 | Operator sends approximate coordinates (lat, lon) | Operator re-localization hint | System uses hint as a high-covariance (~500m) seed for VPR/cross-view re-localization (consumed by Component 5 calibrator), attempts satellite match in new area | threshold_max | position error ≤ 500m initially, ≤ 50m after satellite match | N/A |
-
-### Startup & Handoff
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 23 | System boot with GLOBAL_POSITION_INT available | Normal startup | System reads initial position, initializes Component 5 calibrator state, starts GPS_INPUT output (per AC-NEW-1 cold-start TTFF budget) | threshold_max | GPS_INPUT output begins within 30s of boot (95th percentile) | N/A |
-| 24 | System boot + first satellite match | Startup validation | First satellite match validates initial position, position error drops | threshold_max | position error ≤ 50m after first satellite match | N/A |
-
-### Mid-Flight Reboot Recovery
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 25 | System process killed mid-flight | Companion computer reboot | System recovers: reads FC IMU-extrapolated position, re-initialises Component 5 calibrator state with high uncertainty, loads TRT engines, starts cuVSLAM, performs satellite match | threshold_max | total recovery time ≤ 30s (matches AC-NEW-1 TTFF) | N/A |
-| 26 | Post-reboot first satellite match | Recovery validation | Position accuracy restored after first satellite match | threshold_max | position error ≤ 50m after first satellite match | N/A |
-
-### Object Localization
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 27 | POST /objects/locate with pixel_x, pixel_y, gimbal angles, zoom, known UAV position | Object at known ground GPS | Response: `{ lat, lon, alt, accuracy_m, confidence }` with lat/lon matching ground truth | numeric_tolerance | lat/lon within accuracy_m of ground truth (consistent with frame-center accuracy) | N/A |
-| 28 | POST /objects/locate with invalid pixel coordinates | Out-of-frame pixel | HTTP 422 or error response indicating invalid input | exact | HTTP status 422 | N/A |
-
-### Coordinate Transform Chain
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 29 | Known GPS → NED → pixel → GPS round-trip | Coordinate transform validation | Round-trip error < 0.1m | threshold_max | ≤ 0.1m | N/A |
-
-### API & Communication
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 30 | GET /health | Health check endpoint | HTTP 200, JSON with memory_mb, gpu_temp_c, status fields | exact (status code), regex (body) | status == 200, body contains `"status"` | N/A |
-| 31 | POST /sessions | Start session | HTTP 200/201 with session ID | exact | status ∈ {200, 201} | N/A |
-| 32 | GET /sessions/{id}/stream | SSE position stream | SSE events at ~1Hz with fields: type, timestamp, lat, lon, alt, accuracy_h, confidence, vo_status | regex | each event matches SSE schema | N/A |
-| 33 | Unauthenticated request to /sessions | No JWT token | HTTP 401 Unauthorized | exact | status == 401 | N/A |
-
-### Performance Thresholds
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 34 | Single camera frame (6252x4168) | End-to-end processing time | Total pipeline latency < 400ms (capture → GPS coordinate output) | threshold_max | ≤ 400ms | N/A |
-| 35 | 30-minute sustained operation | Memory usage over time | Peak memory < 8GB, no memory leaks (growth < 50MB over 30min) | threshold_max | peak < 8192MB, growth ≤ 50MB | N/A |
-| 36 | 30-minute sustained operation | GPU thermal | SoC junction temperature stays below 80°C (no throttling) | threshold_max | ≤ 80°C | N/A |
-| 37 | cuVSLAM single frame | VO processing time | cuVSLAM inference ≤ 20ms per frame | threshold_max | ≤ 20ms | N/A |
-| 38 | Satellite matching single frame | Inline cross-view matcher time | SP+LG (TRT FP16/INT8) inline-matcher inference ≤ 200ms / pair on Orin Nano Super @ 25W. (LiteSAM is re-loc-fallback only, ≤ 2s budget — out of inline path.) | threshold_max | ≤ 200ms inline; ≤ 2000ms re-loc fallback | N/A |
-| 39 | TRT engine load | Engine initialization time | All TRT engines loaded within 10s total | threshold_max | ≤ 10s | N/A |
-
-### Satellite Tile Management
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 40 | Mission area definition (200km path, ±2km buffer, zoom 18) | Tile storage calculation | Total storage 500-800MB for zoom 18 + zoom 19 flight path | range | [300MB, 1000MB] | N/A |
-| 41 | ESKF position ± 3σ search radius | Tile selection | Tiles covering search area loaded, mosaic assembled, covers at least 500m radius | threshold_min | coverage radius ≥ 500m | N/A |
-
-### TRT Engine Validation
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 42 | LiteSAM PyTorch model → ONNX → TRT FP16 | TRT engine conversion | Engine builds successfully on Jetson Orin Nano Super | exact | exit_code == 0 | N/A |
-| 43 | TRT engine output vs PyTorch reference (same input) | Inference correctness | Max L1 error between TRT and PyTorch output < 0.01 | threshold_max | L1_max < 0.01 | N/A |
-| 44 | LiteSAM MinGRU operations | TRT compatibility check | All MinGRU ops supported in TRT 10.3 (polygraphy inspect) | exact | unsupported_ops == 0 | N/A |
-
-### Telemetry
-
-| # | Input | Input Description | Expected Result | Comparison | Tolerance | Reference File |
-|---|-------|-------------------|-----------------|------------|-----------|---------------|
-| 45 | Normal operation | Telemetry output rate | NAMED_VALUE_FLOAT messages at 1Hz (gps_conf, gps_drift, gps_hacc) | numeric_tolerance | rate: 1Hz ± 0.2Hz | N/A |
-| 46 | VO tracking lost + 3 satellite failures | Re-localization telemetry | STATUSTEXT with RELOC_REQ sent to ground station | regex | message matches `RELOC_REQ:.*` | N/A |
-
-## Expected Result Reference Files
-
-### position_accuracy.csv
-
-Reference file: `expected_results/position_accuracy.csv`
-
-Contains the ground truth GPS coordinate for each frame in the 60-image test sequence (copied from `coordinates.csv`) plus the acceptance thresholds. Test harness computes haversine distance between estimated and ground truth positions, then applies aggregate criteria.
-
-Thresholds applied to the full 60-frame sequence:
-- ≥ 80% of frames: error < 50m
-- ≥ 60% of frames: error < 20m
-- 0% of frames: error > 100m (no single frame exceeds 100m)
-- Cumulative VO drift between satellite anchors: < 100m
+- The still-image set has expected WGS84 centers but no synchronized IMU, attitude, airspeed, altitude, or timestamp stream.
+- The Derkachi fixture has synchronized video, IMU, and GPS trajectory, but no raw camera calibration, lens distortion, exact camera-to-body transform, attitude, or airspeed columns.
+- The still-image sample cadence is slower than the target 3 fps runtime profile; the Derkachi video is 30 fps and must be sampled to target replay cadence for runtime tests.
+- Final production acceptance requires camera calibration and representative datasets with synchronized camera/IMU plus ground-truth trajectory.

_docs/00_problem/input_data/flight_derkachi/README.md

@@ -0,0 +1,42 @@
+# Derkachi Representative Flight Fixture
+
+## Files
+
+| File | Description | Observed Metadata |
+|------|-------------|-------------------|
+| `flight_derkachi.mp4` | Cropped nadir flight footage for replay | H.264, 880 x 720, 30 fps, about 490.07 s |
+| `data_imu.csv` | Flight-controller telemetry trace exported from the tlog | 4,900 rows at 10 Hz from `Time=0.0` to `489.9`; includes `SCALED_IMU2` and `GLOBAL_POSITION_INT` trajectory fields |
+
+## Test Use
+
+Use this fixture for video/telemetry synchronization checks, representative replay smoke tests, VIO hot-path latency, frame-drop accounting, and trajectory comparison against `GLOBAL_POSITION_INT`. The video and telemetry align at exactly three video frames per telemetry row. Camera intrinsics, lens distortion, raw camera resolution, and exact camera-to-body calibration are still unknown, so this fixture is not sufficient by itself for final production camera calibration or satellite-anchor accuracy claims.
+
+For the test recording, the rotating camera was mechanically fixed in a downward/nadir orientation. Treat the MP4 as a cleaned/cropped replay fixture rather than the raw camera feed.
+
+## Derkachi C6 reference seeding (cycle 3 — AZ-777 + Epic AZ-835)
+
+The end-to-end replay pipeline needs the C6 tile cache pre-populated with the satellite imagery that covers this flight. The seed scripts live under `tests/fixtures/derkachi_c6/`:
+
+| Script | Purpose |
+|--------|---------|
+| `tests/fixtures/derkachi_c6/seed_region.py` (AZ-777 Phase 2) | Bbox-driven seed. Calls `POST /api/satellite/request` on the running `satellite-provider` to onboard the Derkachi area (~50.05–50.15 lat, 36.05–36.15 lon, zoom 15–18). Companion to the existing bbox-download workflow. |
+| `tests/fixtures/derkachi_c6/seed_route.py` (AZ-838 / Epic AZ-835 C2) | Route-driven seed. Reads `derkachi.tlog`, extracts a ≤ 10-waypoint corridor via `replay_input.tlog_route.extract_route_from_tlog`, posts it to `satellite-provider`'s Route API, polls until `mapsReady=true`, and verifies coverage via inventory. ~100× more tile-efficient than the bbox path for this clip. |
+| `tests/fixtures/derkachi_c6/bbox.yaml` | Derkachi bbox + zoom levels + license-attribution metadata (Google Maps Platform ToS + "Imagery © Google" attribution string). |
+| `tests/fixtures/derkachi_c6/README.md` | Step-by-step re-seeding instructions when the `satellite-provider` postgres is wiped; license-attribution operators must propagate; pointer to the parent-suite ticket (TBD) for migrating to a true CC-BY satellite source for production. |
+
+Both seed scripts require:
+
+- A running `satellite-provider` reachable at `SATELLITE_PROVIDER_URL` (typically `https://satellite-provider:8080` inside the Jetson compose network).
+- A valid JWT — either `SATELLITE_PROVIDER_API_KEY` env var or `--auto-mint-jwt` (uses `scripts/mint_dev_jwt.py`).
+- `SATELLITE_PROVIDER_TLS_INSECURE=1` if the parent suite is using the self-signed dev cert (development only — production deploys must validate against a CA-issued cert).
+
+The end-to-end orchestrator test `tests/e2e/replay/test_az835_e2e_real_flight.py` (AZ-840) takes only `(derkachi.tlog, flight_derkachi.mp4, khp20s30_factory.json)` and runs the full 7-step pipeline against a populated C6 — see `_docs/02_document/contracts/replay/replay_protocol.md` Invariant 12.b for the orchestration.
+
+### License attribution caveat (cycle 3)
+
+The Jetson `satellite-provider` instance downloads from the **Google Maps satellite layer** (`lyrs=s`), governed by Google Maps Platform Terms of Service. This fixture and the seed scripts are dev/research use only. Production deployment requires either:
+
+- Google Maps Platform licensing review for offline-cache use, OR
+- A parent-suite ticket to switch satellite-provider's upstream to a true CC-BY satellite source (Esri World Imagery, Mapbox satellite, Sentinel-2, etc.).
+
+The "Imagery © Google" attribution string is recorded in the seeded catalog's metadata and must be propagated downstream by any operator workflow that surfaces the imagery.

_docs/00_problem/input_data/flight_derkachi/camera_info.md

@@ -0,0 +1,34 @@
+# Derkachi camera
+
+Camera model: **Topotek KHP20S30**
+Daylight sensor: 1/2.8" CMOS (Sony IMX291-class, 2.13 MP)
+Image resolution: Full HD 1920×1080 @ 30/60 fps
+Lens: 20× optical zoom, f = 4.7 mm – 94 mm
+
+## Calibration
+
+**File**: [`khp20s30_factory.json`](./khp20s30_factory.json)
+**Acquisition method**: `factory_sheet` (AZ-702 — factory-sheet approximation)
+**Assumed zoom setting**: wide-angle (f = 4.7 mm), HFOV ≈ 59.5°
+
+Per-unit checkerboard refinement is **deferred** (no hardware access to the
+Derkachi unit). The factory-sheet calibration is the cheapest reasonable
+starting point. The residual focal-length error is expected to be in the
+**1–3 %** band; at high AGL this may push horizontal position error past the
+AC-3 100 m budget, in which case AZ-699 (T3 real-flight validation) reports
+the honest finding and a follow-up checkerboard task is filed.
+
+### Why factory-sheet (not checkerboard or PnP-from-tlog)
+
+* **Checkerboard**: needs physical access to the airframe + a known-geometry
+  calibration target. Not in scope for AZ-696.
+* **PnP-from-tlog back-computation**: would require a 5-point task in its own
+  right; deferred as an AZ-696 follow-up if the residual budget proves
+  insufficient.
+
+### Replay-test wiring
+
+`tests/e2e/replay/conftest.py::_calibration_path()` prefers this file when
+present and falls back to `tests/fixtures/calibration/adti26.json` otherwise,
+so dev environments that don't carry the calibration file still exercise the
+AC-1 / AC-2 / AC-5 / AC-6 paths.

_docs/00_problem/input_data/flight_derkachi/flight_derkachi.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9acb97042fc648301d73d3c0fe7d80f7e3e2697000c0d33afa8a7b7a74a20005
+size 282207328

_docs/00_problem/input_data/flight_derkachi/khp20s30_factory.json

@@ -0,0 +1,34 @@
+{
+  "camera_id": "khp20s30_factory",
+  "intrinsics_3x3": [
+    [1680.4469, 0.0,       960.0],
+    [0.0,       1680.4469, 540.0],
+    [0.0,       0.0,         1.0]
+  ],
+  "distortion": [0.0, 0.0, 0.0, 0.0, 0.0],
+  "body_to_camera_se3": [
+    [1.0, 0.0, 0.0, 0.0],
+    [0.0, 1.0, 0.0, 0.0],
+    [0.0, 0.0, 1.0, 0.0],
+    [0.0, 0.0, 0.0, 1.0]
+  ],
+  "acquisition_method": "factory_sheet",
+  "metadata": {
+    "model": "Topotek KHP20S30",
+    "sensor": "1/2.8\" CMOS (Sony IMX291-class), 2.13 MP",
+    "image_resolution_px": [1920, 1080],
+    "sensor_width_mm": 5.37,
+    "sensor_height_mm": 3.02,
+    "assumed_focal_length_mm": 4.7,
+    "focal_length_range_mm": [4.7, 94.0],
+    "assumed_zoom": "wide-angle (max FOV, f=4.7 mm)",
+    "computed_hfov_deg": 59.48,
+    "computed_vfov_deg": 35.62,
+    "intrinsics_formula": "fx = fy = focal_mm * (image_width_px / sensor_width_mm); cx = width/2; cy = height/2",
+    "body_to_camera_convention": "identity-down (nadir, camera-z aligned with aircraft body-z = down per FRD body frame)",
+    "residual_budget_pct": 3.0,
+    "note": "Factory-sheet approximation per AZ-702. The KHP20S30 is a 20x optical-zoom camera (f=4.7-94 mm); the wide-angle f=4.7 mm setting is assumed without per-flight EXIF confirmation. Per-unit checkerboard refinement is deferred — see _docs/00_problem/input_data/flight_derkachi/camera_info.md and the AZ-696 epic. AC-3 (<= 100 m horizontal error) may honestly fail if the assumed focal length is wrong by enough to swamp the 100 m budget at the Derkachi AGL band.",
+    "task": "AZ-702",
+    "epic": "AZ-696"
+  }
+}

_docs/00_problem/problem.md

@@ -1,2 +1,2 @@
-We have a wing-type UAV with a camera pointing downwards that can take photos 3 times per second with a resolution 6200*4100. Also plane has flight controller with IMU. During the plane flight, we know GPS coordinates initially. During the flight, GPS could be disabled or spoofed. We need to determine the GPS of the centers of the next frame from the camera. And also the coordinates of the center of any object in these photos. We can use an external satellite provider for ground checks on the existing photos. So, before the flight, UAV's operator should upload the satellite photos to the plane's companion PC. 
-The real world examples are in input_data folder, but the distance between each photo is way bigger than it will be from a real plane. On that particular example photos were taken 1 photo per 2-3 seconds. But in real-world scenario frames would appear within the interval no more than 500ms. We also don't have IMU data for the test. For now we have to search for the public data for that in internet. We've tried to record that with Mavic 3 Pro Mini, but failed, cause of the closed system if DJI.
+We have a wing-type UAV with a fixed downward navigation camera that can take photos 3 times per second. The authoritative navigation-camera spec is defined in `restrictions.md` as the ADTi 20MP 20L V1, APS-C sensor, about 5472 x 3648 px; older higher-resolution references are superseded. Also plane has flight controller with IMU. During the plane flight, we know GPS coordinates initially. During the flight, GPS could be disabled or spoofed. We need to determine the GPS of the centers of the next frame from the camera. And also the coordinates of the center of any object in these photos. We can use an external satellite provider for ground checks on the existing photos. So, before the flight, UAV's operator should upload the satellite photos to the plane's companion PC. 
+The real world examples are in input_data folder, but the original still-image set has a much larger distance between photos than the target aircraft will have. On that particular example photos were taken 1 photo per 2-3 seconds. But in real-world scenario frames would appear within the interval no more than 500ms. Additional representative data is available in `input_data/flight_derkachi/`: cropped nadir flight footage plus synchronized `SCALED_IMU2` and `GLOBAL_POSITION_INT` telemetry. This supports video/telemetry synchronization, replay, latency, VIO smoke tests, and trajectory comparison against the tlog GPS path. Camera intrinsics, lens distortion, raw camera feed parameters, and exact camera-to-body calibration are still pending, so final production accuracy claims remain gated on calibration data or a separately surveyed representative dataset.

_docs/00_problem/restrictions.md

@@ -1,58 +1,47 @@
 # Restrictions
 
-> **Last revised**: 2026-04-26 (post Mode B Solution Assessment + user-driven addendum on camera spec & zoom level).
+> Last revised 2026-05-07 (cleanup pass — design-independent, IEEE-830 style; only external dependencies, environmental constraints, integration boundaries).
+> Subsequent revision 2026-05-07 (post-SQ6 research): the FC-facing communication protocol entries below were corrected — iNav firmware (master, post-9.0) has no inbound MAVLink external-positioning handler; the project must use a per-FC adapter (MAVLink `GPS_INPUT` for ArduPilot Plane; MSP2 `MSP2_SENSOR_GPS` for iNav). Rationale and L1 sources in `_docs/00_research/02_fact_cards/SQ6_fc_external_positioning.md` / `_docs/00_research/01_source_registry/SQ6_external_positioning.md` Sources #4, #9, #10, #12, #13.
 
 ## UAV & Flight
-
-- Photos are taken by airplane (fixed-wing) type UAVs only.
-- Photos are taken by the navigation camera pointing downwards and fixed (not gimbal-stabilized).
-- Operational area is the eastern and southern parts of Ukraine (east/left of the Dnipro River).
-- Mission profile: 8-hour flights at ~60 km/h cruise. Two route shapes coexist:
-  - **Sector**: up to **10 × 15 km = 150 km²** of dense coverage.
-  - **Transit corridor**: ~**50 km × 1 km = 50 km²** strip in/out of the sector.
-  - **Total operational area: up to ~400 km²** of pre-cached satellite imagery per mission. Cache is **persistent across flights** (not redownloaded each mission). Storage budget **~10 GB** for the satellite tile cache; see AC-NEW-3 for flight-data-recorder budget.
-- Altitude: pre-defined, **≤1 km AGL**. Terrain is assumed flat (operational area is rolling steppe / agricultural land); height differences are negligible.
-- Weather: predominantly sunny daytime operations.
-- Sharp turns occur but are the exception, not the rule. Two consecutive photos may share <5% overlap during a turn (see AC-3.2).
-- **No photo-count cap.** The previously stated "up to 3000 photos per flight" was a legacy operator number from a Mavic-class workflow; it is dropped because (a) it is inconsistent with 8 h × 3 fps, and (b) the system does **not store raw photos at all** (see AC-8.5). Storage is bounded by the tile-cache + FDR caps (~10 GB persistent + 64 GB / flight, AC-NEW-3).
+- Fixed-wing UAVs only; navigation camera fixed downward (no gimbal).
+- Operational area: eastern/southern Ukraine (east of Dnipro).
+- Mission profile: 8-hour flights, ~60 km/h cruise. Sector ≤150 km² + transit corridor ~50 km². Total cached area ≤~400 km², persistent across flights.
+- Altitude ≤1 km AGL; terrain assumed flat (rolling steppe / agricultural).
+- Weather: predominantly sunny daytime; validation must cover seasonal/visibility classes (summer crops, autumn/winter bare fields, cloud/haze, snow if winter, low-texture repetition).
+- Sharp turns are exceptions; consecutive photos may share <5% overlap (AC-3.2).
+- No raw-photo storage (AC-8.5); storage bounded by tile cache + per-flight FDR (AC-NEW-3).
 
 ## Cameras
-
-- The UAV carries **two cameras**:
-  1. **Navigation camera** — fixed, downward-pointing, not autostabilized. Consumed by the GPS-Denied system for position estimation.
-  2. **AI camera** — main mission camera with operator-controllable gimbal angle and zoom. Consumed by onboard AI detection systems.
-- **Navigation camera**: **ADTi 20MP 20L V1, APS-C sensor, ~5472 × 3648 px (≈20 MP)**. APS-C sensor (~23.6 × 15.7 mm). Lens TBD — selected during solution-draft phase to land GSD in the **10–20 cm/px band at 1 km AGL** (drives a frame ground footprint of ~470 m × 314 m to ~980 m × 655 m depending on focal length). Other intrinsics (focal length, exact sensor dimensions, distortion coefficients) are pinned at module-selection time and used by Component-1b orthorectification (pre-flight checkerboard calibration, F-F2).
-- **AI camera pose information available to the GPS-Denied system**: gimbal angle and zoom only. The UAV's instantaneous bank/pitch is **not** published from the autopilot to the AI-camera reasoning path. Object-localization accuracy is therefore scoped to level flight (AC-7.1).
-- Cameras connect to the companion computer over USB, MIPI-CSI, or GigE (specific interface TBD at solution-draft phase, dependent on chosen module).
+- **Navigation camera (pinned)**: ADTi 20MP 20L V1, APS-C ~23.6 × 15.7 mm, ~5472 × 3648 px (≈20 MP). Lens chosen so GSD lands in 10–20 cm/px @ 1 km AGL (frame footprint ~470×314 m to ~980×655 m). Intrinsics + camera-to-body calibration must be obtained pre-flight (e.g., checkerboard).
+- **AI camera**: operator-controlled gimbal angle + zoom (consumed by AI detection systems). The GPS-Denied system supports object localization (AC-7.x) using gimbal angle + zoom only — UAV bank/pitch is not published to that path; AI-camera object localization is therefore scoped to level flight (AC-7.1).
+- Camera-to-companion interface: USB / MIPI-CSI / GigE (lens-module dependent).
 
 ## Satellite Imagery
-
-- **Source: Azaion Suite Satellite Service** (a separate component of the wider Suite). The onboard system is a **consumer** of this service, not a direct customer of commercial providers. Upstream sourcing (Maxar / Airbus / partner agencies / commissioned tasking) is the Satellite Service's concern, not this build's.
-- **Onboard interface to the Service is offline-only**: the companion computer holds a local cache populated **before flight** by syncing from the Service for the operational area (AC-8.3). No satellite imagery is fetched in-flight from the Service.
-- **Mid-flight tile generation (AC-8.4)**: during the mission the companion computer generates fresh tiles from the navigation camera, orthorectified into the basemap projection, deduplicated against the existing cache, and stored locally. On landing, those new tiles are uploaded back to the Suite Satellite Service for ingestion, so the next mission's cache is refreshed by the previous flight.
-- **No raw photo storage** (AC-8.5): the tile is the unit of persistence. Raw nav-camera and AI-camera frames are not retained (except a low-rate failure-thumbnail log for forensics).
-- **Resolution at the cache interface**: 0.5 m/pixel minimum, 0.3 m/pixel ideal (AC-8.1). The architecture is provider-agnostic at the cache boundary; whatever the Suite Satellite Service supplies must meet that bar.
-- **Storage tile zoom level**: **slippy-XYZ z=20 (~30 cm/px, 512×512)** — pinned because the matcher (Component 3) needs ≤~4× scale ratio between the UAV frame (~12 cm/px GSD at 1 km AGL with the 20 MP APS-C camera) and the reference; z=20 gives a 2.5× ratio (workable), z=18 gives a 10× ratio (matcher accuracy breaks down). Storage budget at z=20 across the 400 km² operational area = ~2.8 GB cache + ~30 MB DEM + ~16 MB VPR chunk index ≈ ~3 GB total — well inside the 10 GB cache budget. **VPR retrieval unit is decoupled from the storage tile** (see AC-8.6 below): VPR chunks are derived from the z=20 tile cache at ground-footprint scale (~600–800 m chunks with 40–50 % overlap), independent of the storage zoom level.
-- **Freshness gates** (AC-8.2 / AC-NEW-6) are enforced at runtime: tiles older than 6 months (active-conflict sectors) or 12 months (stable rear sectors) are rejected or down-confidence-weighted. Tiles generated mid-flight are timestamped with the current flight date and treated as fresh.
-- **Free public imagery (Sentinel-2 etc.)** is not on the runtime path. If the Suite Satellite Service ever returns Sentinel-class tiles, the cache rejects them as below the 0.5 m/px floor.
+- **Source**: Azaion Suite Satellite Service (separate Suite component). Onboard system is a consumer; upstream sourcing is the Service's concern.
+- **Onboard interface is offline-only**: companion holds a local cache populated pre-flight from the Service for the operational area (AC-8.3). No in-flight Service calls.
+- **Mid-flight tile generation (AC-8.4)**: companion orthorectifies nav-camera frames into basemap-projected tiles, deduplicates, stores locally; uploads on landing.
+- **Storage policy**: tile is the unit of persistence; no raw frames retained (AC-8.5).
+- **Resolution at cache interface**: ≥0.5 m/px, ideally 0.3 m/px (AC-8.1).
+- **Tile manifest schema**: CRS, tile matrix, dimension, lat-adjusted m/px, capture date, source, compression. Slippy/XYZ zoom (if used) is a provider convention, not a resolution proof.
+- **Cache budget**: 10 GB persistent across the ~400 km² area, including manifests, overviews, and any precomputed indices unless the solution carves out a separate descriptor budget.
+- **Freshness**: enforced per AC-8.2 / AC-NEW-6 (6-month active-conflict / 12-month rear). Mid-flight tiles timestamped current and treated as fresh.
+- **Sentinel-2 / free public imagery**: not on runtime path; cache rejects below the 0.5 m/px floor.
 
 ## Onboard Hardware
-
-- Processing platform: **Jetson Orin Nano Super** — 67 TOPS sparse INT8, **8 GB shared LPDDR5** (CPU and GPU share the same memory pool), **25 W TDP**.
-- Companion computer runs JetPack (Ubuntu-based) with CUDA / TensorRT available.
-- Sustained GPU load may cause thermal throttling; the processing pipeline must stay within the thermal envelope. The cooling solution shall sustain the 25 W power mode for the 8-hour duty cycle at the upper environmental-envelope temperature (AC-NEW-5).
-- Onboard non-volatile storage: budget at least the satellite-cache (~10 GB) **plus** the flight-data-recorder cap (64 GB / flight, AC-NEW-3). Reuse-across-flights tile cache stays resident; per-flight FDR rolls over after cap.
+- **Companion computer (pinned)**: Jetson Orin Nano Super — 67 TOPS sparse INT8, 8 GB shared LPDDR5, 25 W TDP. JetPack (Ubuntu) with CUDA / TensorRT.
+- Cooling sized for 25 W continuous over 8 h at the upper environmental temp (AC-NEW-5).
+- Storage budget ≥ tile cache (~10 GB) + per-flight FDR (64 GB, AC-NEW-3).
 
 ## Sensors & Integration
-
-- High-rate **IMU** data is available from the flight controller via MAVLink.
-- The system communicates with the flight controller via MAVLink. Telemetry plumbing uses **MAVSDK**; the `GPS_INPUT` injection path is implemented via **pymavlink**, since MAVSDK does not expose a native `GPS_INPUT` API.
-- **Autopilot target: ArduPilot only** (with `GPS1_TYPE=14` for MAVLink GPS injection). PX4 is out of scope for the build; if it ever returns to scope it will use `VISION_POSITION_ESTIMATE`, not `GPS_INPUT`. (See `_docs/00_research/00_ac_assessment.md` Q-1.)
-- The system outputs WGS84 GPS coordinates to the flight controller as a replacement for the real GPS module (MAVLink GPS_INPUT, AC-4.3).
-- **Ground station: QGroundControl** is the supported GCS. Mission Planner is not in scope. Telemetry link is bandwidth-limited and is not the primary output channel; per-frame data stays on the local FDR (AC-NEW-3), GCS sees a 1–2 Hz downsampled summary (AC-6.1).
+- **High-rate IMU** available from FC via MAVLink (both ArduPilot Plane and iNav expose IMU telemetry over MAVLink outbound).
+- **Communication protocol (pinned)**: MAVLink for the GCS link (QGroundControl). Companion ↔ FC interface is per-FC: MAVLink for ArduPilot Plane (inbound external positioning + outbound telemetry); MSP2 for iNav (inbound external positioning via `MSP2_SENSOR_GPS`); MAVLink outbound from iNav for telemetry to the GCS is preserved.
+- **Supported flight controllers**: ArduPilot Plane, iNav. PX4 out of scope.
+- **Output to FC**: WGS84 GPS coordinates as a real-GPS replacement, via each supported FC's documented external-positioning interface — MAVLink `GPS_INPUT` for ArduPilot Plane, MSP2 `MSP2_SENSOR_GPS` for iNav (companion is the sole GPS source on iNav; iNav has no dual-GPS arbitration). Per-FC parameter wiring (EKF source-set on ArduPilot; GPS provider/UART selection on iNav) and source-label out-of-band channel are design choices; outcome contract is AC-4.3.
+- **Ground station**: QGroundControl (Mission Planner out of scope). Telemetry link bandwidth-limited; per-frame data stays on local FDR (AC-NEW-3); GCS sees 1–2 Hz downsampled summary (AC-6.1).
+- **Representative data**: see `input_data/` (still images), `input_data/flight_derkachi/` (cropped nadir video + synchronized `SCALED_IMU2` + `GLOBAL_POSITION_INT`). Production acceptance still requires camera intrinsics, distortion, camera-to-body calibration, and synchronized representative flight data (frames + FC IMU/attitude/airspeed/altitude + emitted MAVLink + ground-truth trajectory).
 
 ## Failsafe & Safety
-
-- If the GPS-Denied system fails to produce any position estimate for **>3 s**, the autopilot falls back to IMU-only dead reckoning (AC-5.2). N=3 s rides through one sharp turn at cruise speed without tripping the failsafe.
-- The system must satisfy the false-position safety budget in AC-NEW-4 (P(error >500 m) <0.1%, P(error >1 km) <0.01% per flight).
-- Cold-start time-to-first-fix budget is **<30 s** from companion-computer boot (AC-NEW-1); spoofing-promotion latency is **<3 s** from FC's GPS-loss signal (AC-NEW-2).
+- If no estimate produced for >3 s → autopilot falls back to IMU-only dead reckoning (AC-5.2). 3 s rides through one sharp turn at cruise.
+- False-position safety budget: AC-NEW-4 (P(>500 m) <0.1 %, P(>1 km) <0.01 % per flight).
+- Cold-start TTFF <30 s (AC-NEW-1); spoofing-promotion latency <3 s (AC-NEW-2).

_docs/00_research/00_question_decomposition.md

@@ -0,0 +1,253 @@
+# Question Decomposition
+
+> Mode A Phase 2 (Initial Research — Problem & Solution Draft).
+> Phase 1 (AC & Restrictions Assessment) was skipped per user decision after a cleanup pass that stripped implementation details from `acceptance_criteria.md` and `restrictions.md` (commit `12cc5a4`); AC/restrictions are treated as fixed inputs.
+
+## Original Question
+
+Design the GPS-denied onboard navigation system for a fixed-wing UAV operating in eastern/southern Ukraine, satisfying every AC in `_docs/00_problem/acceptance_criteria.md` under the constraints in `_docs/00_problem/restrictions.md`. Recommend a concrete component-by-component architecture and tech stack.
+
+## Research Output Class
+
+**Technical-component selection.** All technical-component gates apply (per-mode API capability verification, Component Applicability Gate, Restrictions × Candidate-Mode sub-matrix, MVE evidence, mandatory `context7` lookups for every lead library/SDK candidate).
+
+## Question Type
+
+**Decision Support** (per Mode A Phase 2 default). Sub-flavour: multi-component decision support — weighing trade-offs across ~10 interlocking component areas under hard real-time + memory + safety budgets.
+
+## Project Context Summary (from `_docs/00_problem/`)
+
+- **What is being built**: an onboard companion-PC system that replaces real GPS for a fixed-wing UAV when GPS is denied/spoofed, by combining nav-camera frames + FC IMU + a pre-cached satellite tile basemap, and emits standard MAVLink external-positioning messages to ArduPilot or iNav at frame rate.
+- **Operating area**: eastern/southern Ukraine, active-conflict zones (war-zone scene change is a first-class concern, not an edge case).
+- **Mission profile**: 8-hour fixed-wing flights, ~60 km/h cruise, ≤1 km AGL, ~400 km² operational area.
+- **Pinned external deps**: ADTi 20MP 20L V1 nav camera (APS-C); Jetson Orin Nano Super 8 GB / 25 W; MAVLink protocol; ArduPilot + iNav as supported FCs; QGroundControl as GCS; Azaion Suite Satellite Service (offline cache interface ≥0.5 m/px).
+- **Hard runtime envelope**: <400 ms p95 end-to-end latency (camera → MAVLink), <8 GB shared CPU+GPU RAM, 25 W TDP at +50 °C ambient for 8 h continuous, no in-flight network, 10 GB persistent tile cache + 64 GB per-flight FDR.
+- **Hard safety envelope**: P(error >500 m) <0.1 % per flight, P(error >1 km) <0.01 % per flight; honest covariance reporting; explicit `dead_reckoned` failsafe under simultaneous GPS spoof + visual blackout; cache-poisoning probability bounds for tiles written back to the Service.
+
+## Project Constraint Matrix
+
+| Dimension | Binding constraint |
+|---|---|
+| **Inputs available** | Nav camera frames @ 3 fps (5472×3648, ~12 cm/px GSD @ 1 km AGL); FC IMU (high rate via MAVLink); FC attitude/airspeed/altitude; pre-cached satellite tiles ≥0.5 m/px (offline); operator re-loc hint via GCS (rare). |
+| **Outputs required** | WGS84 position to FC via MAVLink external-positioning message(s) accepted by ArduPilot AND iNav; per-frame estimate carrying honest 95 % covariance, source label `{satellite_anchored, visual_propagated, dead_reckoned}`, and `last_satellite_anchor_age_ms`; mid-flight ortho-tiles written to local cache with quality metadata; 1–2 Hz GCS summary; FDR records per AC-NEW-3. |
+| **Hardware fixed** | Jetson Orin Nano Super (67 TOPS sparse INT8, 8 GB shared LPDDR5, 25 W TDP, JetPack/CUDA/TensorRT). |
+| **Lifecycle** | Real-time embedded; offline (no in-flight network); 8 h continuous; persistent tile cache across flights; FDR rollover. |
+| **Non-functional** | <400 ms p95 latency; <8 GB shared RAM; ≤25 W power at +50 °C ambient; AC-1.1/1.2 accuracy; AC-2.1/2.2 registration & MRE; AC-3.x resilience; AC-NEW-1 cold-start <30 s; AC-NEW-2 spoof promotion <3 s; AC-NEW-4 false-position safety; AC-NEW-7 cache-poisoning safety; AC-NEW-8 blackout failsafe. |
+| **Hard disqualifiers** | Anything requiring >8 GB RAM peak (CPU+GPU shared); anything not runnable under JetPack on Orin Nano Super; anything requiring in-flight cloud calls; anything that cannot honestly report covariance; anything that does not have a runnable example for monocular nadir UAV input over season-matched satellite tiles; anything whose license blocks military / dual-use deployment. |
+
+## Research Subject Boundary
+
+| Dimension | Boundary |
+|---|---|
+| **Population** | Fixed-wing UAVs, downward-fixed monocular nav camera, Jetson-class edge HW, ArduPilot or iNav autopilot. Excludes: multirotors, gimbal-stabilised nav cams, server/cloud GPS-denied stacks, PX4 (out of scope), commercial sat-imagery direct integration (Service handles upstream). |
+| **Geography** | Eastern/southern Ukraine — agricultural steppe, active-conflict scene change. Validation must include this geography or representative analogues (low-texture cropland, snow, war-zone destruction). |
+| **Timeframe** | Production deployment 2026; tools / libraries / models considered must be currently maintained (commits/releases in last 18 months OR explicit long-term-stable status). Critical-novelty domain — see Step 0.5 timeliness assessment. |
+| **Operating context** | Real-time embedded; offline in-flight; 8 h continuous duty; 25 W power envelope; 8 GB shared CPU+GPU memory; thermal envelope to +50 °C ambient. |
+| **Required interfaces** | Inputs: ADTi 20MP nav cam, FC IMU (MAVLink), satellite tile cache. Outputs: MAVLink external-positioning to ArduPilot AND iNav; QGroundControl summary; FDR; tile write-back to Suite Service on landing. |
+| **Non-functional envelope** | Per AC-1 to AC-8 plus AC-NEW-1 to AC-NEW-8. Hardest binding constraints: 400 ms p95 end-to-end; 8 GB shared RAM; AC-NEW-4 false-position probability bounds; AC-NEW-7 cache-poisoning probability bounds. |
+
+## Sub-Questions
+
+| ID | Sub-question |
+|---|---|
+| SQ1 | What existing/competitor GPS-denied UAV navigation systems exist (academic + open-source + commercial + military), and which of them have been validated on fixed-wing UAVs in active-conflict environments? What works, what fails? |
+| SQ2 | What is the canonical decomposition of "monocular nadir UAV ↔ pre-cached satellite basemap localization" into pipeline components? Is the decomposition below complete, or are there industry-standard components missing? |
+| SQ3 | For each component (VO/VIO, VPR, cross-domain registration, single-frame orthorectification, sensor-fusion estimator, tile cache + spatial index, on-Jetson inference runtime, MAVLink FC adapter, dataset/SITL validation infrastructure): what option families exist (simple baseline / production / open-source / commercial / SOTA / adjacent-domain / no-build), and what are the leading candidates as of 2026? |
+| SQ4 | For each lead candidate per component: what are the documented runtime/memory/latency/license/maintenance constraints, and how do they bind against the Project Constraint Matrix? Per-mode API capability verification with `context7` for every library/SDK lead. |
+| SQ5 | What are the documented failure modes and real-world deployment lessons for each component family? In particular: VPR collapse under cropland repetition, DINOv2/foundation-model cost on Jetson at int8, RANSAC degeneracy at sharp turns / low texture, EKF over-confidence on cross-domain matches, ortho geometric error from unknown bank/pitch. |
+| SQ6 | How do **ArduPilot Plane** (current stable) and **iNav** (current stable) each accept external positioning input via MAVLink? What message types does each support? Where do their interfaces diverge, and what is the documented status of each interface (stable / experimental / known bugs)? |
+| SQ7 | What public datasets, benchmarks, and SITL/replay environments exist for cross-validating monocular nadir UAV navigation against satellite basemaps in season-matched + change-affected conditions? AerialVL, AerialExtreMatch, others? |
+| SQ8 | What are the security and safety considerations specific to the AC-NEW-4 (false-position) and AC-NEW-7 (cache-poisoning) safety budgets, including spoofing-detection signals from FC, ortho-tile geo-alignment quality estimation, and write-back cache-poisoning controls? |
+| SQ9 | What does the system look like end-to-end — wiring, scheduling, threading model, inference scheduling on shared CPU+GPU memory, cold-start sequencing, FDR rotation, and pre-flight cache provisioning workflow? (synthesis question, answered in Step 8) |
+
+## Component Areas (search plan)
+
+For each component below, the search plan covers all option families per `Component Option Search Plan` rules (`research/steps/03_engine-investigation.md` → "Component Option Breadth").
+
+| # | Component area | Required outputs | Key option families to enumerate |
+|---|----------------|------------------|----------------------------------|
+| C1 | **Visual / Visual-Inertial Odometry** (frame-to-frame motion when satellite anchor is unavailable) | Relative 6-DoF pose between consecutive frames or short windows; output frequency ≥3 Hz; metric scale (with IMU) | Classical (VINS-Mono / VINS-Fusion / OpenVINS), Kimera, ORB-SLAM3, OKVIS2, MSCKF-class, learning-based (DROID-SLAM, DPVO), pure VO baseline (KLT + RANSAC homography), no-build (skip and rely on pure satellite re-anchor every frame) |
+| C2 | **Visual Place Recognition (VPR)** — UAV nadir frame → top-K satellite chunks | Compact global descriptor per UAV frame and per cache chunk; cosine-rank top-K candidates | NetVLAD class, MixVPR, EigenPlaces, BoQ, AnyLoc (DINOv2 + VLAD), CricaVPR, foundation-model direct retrieval (DINOv2/DINOv3/SAM 2 / SuperGlobal) |
+| C3 | **Cross-domain registration** (UAV nadir ↔ ortho satellite tile, after VPR top-K) | Sub-pixel alignment + 6-DoF camera pose w.r.t. tile, with inlier ratio + covariance | Local-feature matching (SuperPoint+SuperGlue / LightGlue / DISK+LightGlue / ALIKED+LightGlue / XFeat), dense matchers (LoFTR / RoMa / DKM / MASt3R), classical (SIFT+RANSAC), specialized cross-domain (CMRNet+, CroCoMatch class), templating (mutual-information / ECC), no-build (skip cross-domain; rely on direct frame-to-tile homography from VPR retrieval) |
+| C4 | **Single-frame orthorectification** (nav frame → basemap-aligned tile, given current pose) | Ortho-rectified tile chunk with geo metadata + quality score | Single-frame perspective warp with flat-earth assumption; OpenCV homography; bundled-DEM-aware (rare for flat steppe — likely overkill); GDAL warp utilities; custom GPU shader on Jetson |
+| C5 | **State estimator / sensor fusion** (VO + IMU + sat anchors → fused estimate with covariance) | WGS84 position + 3D velocity + attitude + 6×6 covariance, frame-rate output, honest covariance, source label | EKF (manual), ESKF (manual or via library), MSCKF, factor-graph (GTSAM, iSAM2), particle filter, learned (out-of-scope for safety budget). Supporting: Mahalanobis outlier gates |
+| C6 | **Tile cache + spatial index** (storage + retrieval of basemap tiles + descriptors, with manifests, freshness, dedup, and write-back) | mmap-friendly storage; ANN over global descriptors; spatial query for geographic prior; manifest schema per AC | Storage: GeoTIFF + COG, MBTiles, custom flat layout. ANN: FAISS (IVF/PQ/HNSW), hnswlib, ScaNN, brute-force (small index). Spatial: R-tree / KD-tree / GeoPandas / SQLite+SpatiaLite. Manifest: SQLite, JSON-per-tile, Parquet sidecar |
+| C7 | **On-Jetson inference runtime** | INT8/FP16 inference of the chosen VPR + matcher models within latency + memory budget | TensorRT (native), Torch-TensorRT, ONNX Runtime + TRT EP, NVIDIA Triton (probably overkill), pure PyTorch fp16, NVIDIA DeepStream (for video), CUDA-Python custom kernels |
+| C8 | **MAVLink FC adapter** (per-FC external-positioning emission + spoofing-signal subscription, for ArduPilot AND iNav) | MAVLink frames consumed by ArduPilot Plane and iNav as external position; spoofing signals consumed from each FC | Libraries: `pymavlink` (per-message), MAVSDK (high-level), ArduPilot/iNav SITL for verification. Per-FC choice of message: `GPS_INPUT` vs `ODOMETRY` vs `VISION_POSITION_ESTIMATE` vs `GLOBAL_POSITION_INT` (documented capability per FC must be verified, not assumed) |
+| ~~C9~~ | ~~**Datasets + SITL / replay**~~ — **DROPPED from research scope per 2026-05-08 restructure (user choice A)**; deferred to **Test Spec (greenfield Step 5)**. See "C9 / SQ7 Restructure" section below. | — | — |
+| C10 | **Pre-flight cache provisioning + sector classification + freshness pipeline** (RESEARCH SCOPE NARROWED 2026-05-08 to cross-coupling minimal — see "C10 Scope Restructure" section below) | (in research scope) confirmed orchestration mechanism for descriptor-cache rebuild (D-C6-3) + TensorRT engine build (D-C7-7) at pre-flight; on-disk artifact format(s); time/memory budget; failure-mode + retry behavior. (deferred to Plan-phase) operator CLI/desktop tool design, sector classification heuristics, freshness pipeline workflow. | (in research scope) FAISS Python API for write_index/read_index orchestration; TensorRT build orchestration `trtexec` CLI vs Python `IBuilderConfig` vs Polygraphy. (deferred) custom CLI/desktop, QGC plan files, MAVProxy, Mission Planner integration patterns. |
+
+## Perspectives Chosen (≥3 mandatory)
+
+1. **Implementer / Engineer** — Will the chosen stack actually compile, link, and run on the pinned Jetson within the latency + memory budget? Pitfalls of MAVLink GPS injection on each FC. Sub-pixel registration on UAV-nadir × ortho satellite. Inference-scheduler contention on shared CPU+GPU memory.
+2. **Practitioner / Field** — What do UAV teams actually report from GPS-denied missions in real war-zone deployments? (Ukraine context if findable; otherwise analogous high-stakes deployments.) Real-world VPR collapse on agricultural cropland / snow / season change. Real-world FDR usefulness for post-mission forensics.
+3. **Domain expert / Academic** — Recent (2024–2026) VPR + cross-domain matching benchmarks and their relative ranks under cross-season / cross-domain / cross-altitude conditions. Foundation-model-based VPR (AnyLoc, BoQ, MASt3R) — academic claims vs reproducibility. Recent factor-graph vs ESKF comparisons.
+4. **Contrarian / Devil's advocate** — Why might foundation-model VPR fail on the Jetson budget? Where does cross-domain matching degrade silently? When does ortho-tile write-back amplify bad poses? When does honest covariance turn into "system never trusts itself" (over-cautious failure)?
+
+## Search Query Variants per Sub-Question
+
+(Detailed query lists are appended below per sub-question; these will be executed in Step 2 and saved to the `01_source_registry/` folder, indexed by `01_source_registry/00_summary.md`. The shape is shown here so the search plan is auditable; the full execution log will populate downstream files.)
+
+**SQ1** (existing systems / competitors): "GPS-denied UAV navigation 2025", "visual GPS denied fixed wing UAV", "satellite map matching UAV localization 2024 2025", "Ukraine UAV GPS spoofing countermeasures", "ARL ANT Project visual navigation", "vision-based GPS replacement UAV production", "UAV GPS spoofing real-world deployment 2025".
+
+**SQ2** (canonical pipeline): "visual aerial localization pipeline survey", "UAV satellite map matching architecture", "monocular UAV global localization pipeline 2024 2025".
+
+**SQ3 / SQ4** (per-component candidates + binding): per-component query templates (5+ variants each) — see Step 2 plan in `01_source_registry/00_summary.md` once initialised. Each lead library/SDK candidate triggers the mandatory `context7` per-mode capability verification per `research/steps/03_engine-investigation.md`.
+
+**SQ5** (failure modes): "VPR cropland failure", "DINOv2 Jetson Orin Nano latency", "SuperGlue LightGlue Jetson Orin", "ESKF cross-domain over-confidence", "RANSAC homography low-texture failure UAV", "ortho photo geometric error airframe tilt".
+
+**SQ6** (ArduPilot vs iNav external positioning): "ArduPilot Plane GPS_INPUT external", "ArduPilot ODOMETRY EKF3 source switching", "iNav external positioning MAVLink GPS_INPUT", "iNav MAVLink GPS substitute", "iNav GPS denied flight 2025", "ArduPilot vs iNav external nav comparison".
+
+**SQ7** (datasets): "AerialVL dataset", "AerialExtreMatch", "VPR-Bench cross-season aerial", "Mid-Air UAV dataset", "Mavic Mavik UAV public flight dataset", "satellite-aerial cross-view localization benchmark".
+
+**SQ8** (safety): "MAVLink GPS_RAW_INT spoofing detection", "EKF lane switch ArduPilot", "covariance under-reporting risk EKF", "geo-misalign detection ortho tile".
+
+## Completeness Audit
+
+Probes (per `references/comparison-frameworks.md` → Decomposition Completeness Probes — applied here without re-reading the full file; will reconcile during Step 2):
+
+| Probe | Coverage |
+|---|---|
+| Functional decomposition complete? | C1–C10 cover all data flows from camera in to MAVLink out + back. ✓ |
+| Non-functional dimensions covered? | Latency, memory, accuracy, safety, freshness, security all in Project Constraint Matrix. ✓ |
+| Failure-mode dimension covered? | SQ5 explicitly. ✓ |
+| Cost / TCO dimension? | Hardware is pinned (Jetson Orin Nano Super); Service-side cost is out of scope; SW cost = mostly open-source candidates. Will revisit during Phase 3 (tech stack consolidation) if commercial options emerge. ✓ |
+| Maintenance / community-health dimension? | SQ4 binds it per candidate. ✓ |
+| Adjacent-domain dimension? | Robot SLAM, AGV warehouse navigation, aerial photogrammetry will be searched as analogues. ✓ |
+| Validation / dataset coverage? | **Deferred to Test Spec (greenfield Step 5) per 2026-05-08 C9 / SQ7 restructure** — fixture-class, not research-class. Dataset shortlist preserved for handoff. |
+| Integration / boundary coverage? | SQ6 (FC adapters) + C8 + C10 (pre-flight provisioning). ✓ |
+| Operational/human-factors? | Pre-flight cache provisioning (C10) and operator re-loc hint (AC-3.4) covered. Mission-planning UX is out of scope. ✓ |
+| Security / threat model? | SQ8. Will deepen in Phase 4 (Security Deep Dive) if invoked. ✓ |
+
+No major gap detected at decomposition time. If domain-discovery searches in Step 2 surface a missed dimension, a "gap-fill" entry will be appended here.
+
+## Notes on Output-Class Mode-Verification
+
+Because this is **Technical-component selection**, every lead library/SDK candidate triggers:
+- Pinned mode/configuration sentence in `02_fact_cards/Cx_*.md` (per-component sub-files).
+- `context7` lookup with the three mandatory queries (mode enumeration; project's exact mode runnable example; disqualifier probe).
+- MVE block per candidate.
+- Per-numbered-Restriction and per-numbered-AC binding (`Pass` / `Fail` / `Verify` / `N/A`).
+- Two modes of one library = two distinct candidates.
+
+## Step 0.5 — Novelty Sensitivity Assessment
+
+**Classification: Critical sensitivity.**
+
+Justification:
+- Foundation-model VPR is moving fast: DINOv2 (Apr 2023), AnyLoc (Aug 2023), BoQ (CVPR 2024), MASt3R (May 2024), MASt3R-SfM / new VPR-leader candidates 2025; rankings on cross-season aerial benchmarks have shifted multiple times since 2023.
+- ArduPilot Plane / iNav external-positioning interfaces have moved: ArduPilot EKF3 source-switching parameters and known double-fusion bugs between `GPS_INPUT` and `ODOMETRY` were a moving target through 2024–2025; iNav GPS-denied support has matured separately.
+- TensorRT / JetPack stacks on Jetson Orin Nano Super have version-dependent INT8 quantisation behaviour and runtime tooling differences worth verifying against current releases.
+- Public aerial-localization datasets (AerialVL, AerialExtreMatch, etc.) have had multiple revisions and added splits.
+
+Source-time-window rules for this run:
+- **Lead-candidate selection / SOTA claims**: prioritise sources from **last 6 months**; allow up to **18 months** if no newer source covers the same claim and the older source is the official authority.
+- **Established baselines / classical algorithms** (KLT, RANSAC, EKF, ORB, SIFT, GTSAM): no time window — canonical references are fine.
+- **Library/SDK API behaviour**: must be verified against the **currently shipped version** at the time of search (`context7` mandatory per lead candidate; release notes / changelog cross-checked).
+- **Cross-validation**: every Critical-sensitivity claim that drives a candidate selection must have **≥2 independent sources** or one official + one runnable MVE; single-source SOTA claims must be downgraded to `Experimental only` at Step 7.5 unless cross-validated.
+
+## SQ2 Closure — Pipeline-component coverage table (Mode A Phase 2, Step 3 result)
+
+The C1–C10 decomposition was sanity-checked against five independent surveys/benchmarks (Skoltech aerial-VPR survey, U.Maine cross-view survey, OrthoLoC benchmark, AnyVisLoc benchmark, NUDT 2026 absolute-VL survey — all logged in `01_source_registry/SQ2_canonical_pipeline.md` as Sources #38–#42). The canonical hierarchical framework `retrieval → matching → pose-estimation` is unanimously confirmed; project's split is **canonical, not novel**. Two augmentations are required.
+
+| Survey/benchmark canonical stage | Project component | Coverage status | Required action |
+|---|---|---|---|
+| Image retrieval (global VPR) | **C2 — VPR** | ✅ covered | None |
+| Re-ranking (top-N inlier-based) | (implicit, inside C2/C3) | ⚠️ implicit | Promote to explicit sub-stage in `solution_draft01` |
+| Local image matching (2D-2D, sparse or dense) | **C3 — Cross-domain registration** | ✅ covered | Add Top-N inlier re-rank requirement |
+| AdHoP-style perspective preconditioning | (not represented) | ❌ missing | Add as optional sub-stage between C3 and C4, gated on Jetson latency budget |
+| 2D-3D lift via DSM | (not represented; current cache is 2D ortho only) | ❌ architectural decision required | **Decision required from user** — see "Open architectural decisions" below |
+| Pose estimation (PnP + RANSAC + LM) | **C4 — Pose estimation** | ✅ covered | None |
+| State estimator / fusion | **C5 — Estimator / fusion** | ✅ covered | Augmented with covariance-honesty contract (already from AC-NEW-4) |
+| IMU + VIO contract | **C1 (VIO)** + **C6 (Tile cache)** ⁂ | ✅ covered | Add yaw σ ≤ 5°, pitch σ ≤ 5° hard contract (Fact #24) |
+| Tile cache + scheduler | **C6 (Tile cache + spatial index)** | ✅ covered | Add 20% covisibility runtime invariant (Fact #27) |
+| On-Jetson runtime | **C7 — On-Jetson inference runtime** | ✅ covered | Pre-screen prunes non-viable candidates (Fact #26) |
+| Anti-spoof / FC adapter | **C8 — MAVLink FC adapter** | ✅ covered | Already addressed by SQ6 |
+| Datasets / SITL / replay | **Deferred to Test Spec (greenfield Step 5)** per 2026-05-08 C9 / SQ7 restructure | ⚠️ moved out of research scope | Test Spec owns dataset-corpus selection, SITL framework choice (ArduPilot Plane SITL + iNav SITL/HITL), and replay framework choice |
+| Pre-flight cache provisioning | **C10 — Pre-flight cache + sector classification** | ✅ covered | None |
+
+⁂ The "IMU integration" concern lives in C1 (VIO) and partially flows from FC IMU; there is no separately numbered IMU component in the original C1–C10 split. SQ2 confirms this was correct — IMU is best owned by C1 (VIO) which already produces the yaw/pitch attitude. The σ ≤ 5° contract belongs on C1's output interface.
+
+### SQ2 — Architectural decisions (resolved by user, 2026-05-07)
+
+| # | Decision | Choice | Implication for SQ3+SQ4 |
+|---|---|---|---|
+| 1 | DSM dependency on Suite Sat Service tile cache (Fact #23) | **(a) 3-DoF acceptance** — fix attitude from IMU/VIO, ignore DSM; current 2D-ortho cache contract preserved. | C6 (Tile cache) candidate matrix excludes DSM-dependent storage formats. C3 (matcher) candidates evaluated on 2D-2D output (homography) only. Yaw/pitch σ ≤ 5° (Fact #24) is **noted as an empirical requirement on C1's output but NOT bound as a hard interface contract** — emerges as an output of C1 candidate selection in SQ3+SQ4. AC-1.1.1 (≤80 m at 1 km AGL) likely satisfied per DSMAC-class lineage in Fact #17; if AC ever tightens, revisit option (b). |
+| 2 | AdHoP refinement loop (Fact #22) | **(b) Conditional** — only invoked when initial reprojection error exceeds a threshold. | C3 (matcher) latency budget = base (single-pass) + AdHoP-conditional overhead (worst-case 2× when triggered). Per-frame Jetson MVE must measure both modes. The reprojection-error threshold becomes a SQ3+SQ4 hyperparameter. |
+| 3 | Top-N re-rank promotion (Fact #25) | **(a) Promote** to an explicit named sub-stage between C2 and C3. | SQ3+SQ4 will hyperparameter-sweep N ∈ {5, 10, 15, 20}; C2 candidates evaluated jointly with re-rank cost. Top-N re-rank by inlier-count is now a hard pipeline component, not implicit. |
+
+### SQ2 — Component-pruning carried into SQ3+SQ4 (Jetson-pre-screen result)
+
+Per Fact #26 (RTX-3090-measured runtime → conservative Jetson-Orin-Nano translation):
+
+- **C2 candidates entering SQ3+SQ4 with mandatory Jetson MVE**: MixVPR, SALAD, SelaVPR, EigenPlaces, NetVLAD.
+- **C2 candidates entering SQ3+SQ4 conditional on INT8 quantization path**: AnyLoc, BoQ, DINOv2-VLAD.
+- **C2 candidates pruned outright**: SuperGlue-as-reranker (latency).
+- **C3 candidates entering SQ3+SQ4 with mandatory Jetson MVE**: LightGlue, XFeat, XFeat*, SP+LightGlue (NGPS template confirmed).
+- **C3 candidates pruned outright**: RoMa, MASt3R, DKM (dense-matcher latency on Jetson).
+- **C3 candidates as "AerialExtreMatch reference points" only**: GIM+DKM, GIM+LightGlue (per Source #40 — accuracy benchmark, not for production deployment).
+
+## C9 / SQ7 Restructure (2026-05-08, user choice A)
+
+**Decision**: drop C9 (Datasets + SITL / replay) entirely from the research scope. Defer dataset-corpus selection, SITL framework choice (ArduPilot Plane SITL + iNav SITL/HITL), and replay framework choice (custom vs PX4-Avionics-Replay-style) to **Test Spec (greenfield Step 5)**. Pull D-C7-1 (calibration-dataset-strategy) back inside C7 batch 1 and close it there.
+
+**Rationale**: datasets are test fixtures, not architectural commitments. They feed into Test Spec → Decompose Tests → Implement Tests, not into the deployed pipeline on the Jetson. They don't bind against the AC-4.1 / AC-4.2 / R-NEW-2 / R-NEW-4 envelope. Choosing among AerialVL S03 vs AerialExtreMatch vs VPR-Bench vs MahalNotchVPR / Mid-Air UAV vs the project's own Mavic + Derkachi flight footage is a "what evidence proves the system meets AC-X" question, not a "what gets implemented on the Orin Nano" question. SITL and replay framework choice are test-infra commitments rather than runtime commitments; SITL framework is largely deterministic at this point (ArduPilot Plane SITL + iNav SITL/HITL are the canonical paths the locked C8 closure already implies).
+
+**Effective changes**:
+- **Component Areas table**: C9 removed; remaining components are C1–C8 + C10.
+- **Sub-Questions table**: SQ7 is deferred to Test Spec (Step 5) — its query variants and dataset shortlist remain documented here for handoff but are not researched in this Mode A run.
+- **SQ2 closure table**: "Datasets / SITL / replay" row → "Deferred to Test Spec".
+- **D-C7-1 (calibration-dataset-strategy)**: closed inside C7 batch 1. Strategy = prefer real UAV nadir flight footage at ~1 km AGL over season-matched satellite tiles as the calibration corpus distribution; specific fixture-file selection (AerialVL S03 vs project's Mavic + Derkachi clips vs other corpora) is fixture-class and delegated to Test Spec. Synthetic-tile augmentation via random homography is a documented low-data fallback, only invoked if real flight footage is insufficient for Recall@K-target calibration.
+- **Cross-component gates**: D-C7-1 is no longer cross-coupled to C9; owner narrows to Plan-phase architect (closed at research time).
+- **Cross-row dependencies in C7 / C8 fact cards and fit-matrix files**: every "C9 datasets / SITL / replay row when opened" reference becomes "Test Spec (Step 5) when opened".
+
+**Carryforward to Test Spec (Step 5)** — preserved here so Test Spec's first invocation has the handoff payload ready:
+- **Dataset shortlist**: AerialVL (VISTA / NTU), AerialExtreMatch, VPR-Bench, MahalNotchVPR / Mid-Air UAV, project's own Mavic + Derkachi flights.
+- **SITL frameworks**: ArduPilot Plane SITL (canonical), iNav SITL/HITL (canonical); Gazebo / Webots noted-and-rejected as overkill for the spoof-promotion + visual-blackout failsafe scenarios that AC-NEW-2 and AC-NEW-8 actually exercise.
+- **Replay frameworks**: PX4-Avionics-Replay-style canonical reference; custom Python harness as the lightweight default if PX4 replay's MAVLink-injection point doesn't cleanly match the C8 closure's per-FC injection cadence (5 Hz GPS_INPUT for AP / 5 Hz MSP2_SENSOR_GPS for iNav).
+- **SQ7 query variants** (carried forward verbatim from above): "AerialVL dataset", "AerialExtreMatch", "VPR-Bench cross-season aerial", "Mid-Air UAV dataset", "Mavic Mavik UAV public flight dataset", "satellite-aerial cross-view localization benchmark".
+- **Test-coverage obligations Test Spec must answer**:
+  - Which corpora exercise which AC (AC-1.1 / AC-1.2 / AC-2.1 / AC-2.2 / AC-3.1 / AC-3.2 / AC-3.3 / AC-3.4 / AC-NEW-1 / AC-NEW-2 / AC-NEW-4 / AC-NEW-7 / AC-NEW-8).
+  - SITL test-harness shape exercising AC-NEW-2 spoof-promotion <3 s end-to-end on **both** ArduPilot Plane SITL **and** iNav SITL/HITL (per locked C8 batch 1 closure cross-component decision D-C8-2).
+  - Replay-fixture format compatible with both C8 injection paths (pymavlink GPS_INPUT for AP, YAMSPy MSP2_SENSOR_GPS for iNav).
+  - INT8 calibration corpus pin (specific files satisfying the C7 batch 1 D-C7-1 strategy = real UAV nadir flight footage at ~1 km AGL over season-matched satellite tiles).
+
+## C10 Scope Restructure (2026-05-08, user choice C — cross-coupling minimal)
+
+**Decision**: narrow C10 (Pre-flight cache provisioning + sector classification + freshness pipeline) research scope to the two cross-coupling confirmation sub-areas. Defer the operator-side CLI/desktop tool, sector classification heuristics, and tile age-stamping/freshness schema to Plan-phase as `operator tooling design` out-of-research-scope.
+
+**In-scope (C10 batch 1)**:
+1. **D-C6-3 confirmation** — descriptor-cache rebuild trigger pipeline. Recommendation inherited from C6 batch 1 (Fact #92 + D-C6-3) = `periodic rebuild during C10 pre-flight provisioning + faiss.write_index serialize + load-at-takeoff in <5 s`. Confirmation work: pin the orchestration tool (FAISS Python API vs subprocess invocation), the trigger semantics (manifest hash change vs operator-manual vs new-tile-delivered), the on-disk file format, the rebuild time budget at pre-flight, and the failure-mode + retry behavior.
+2. **D-C7-7 confirmation** — TensorRT engine-build pipeline. Recommendation inherited from C7 batch 1 (Fact #94 + D-C7-7) = `primary build-on-deployed-Jetson during pre-flight + reference-Jetson-built engines as fallback`. Confirmation work: pin the build-orchestration tool (`trtexec` CLI vs Python `IBuilderConfig` vs Polygraphy), the calibration-corpus shipping mechanism into the pre-flight build (per D-C7-1 closure: real UAV nadir flight footage at ~1 km AGL over season-matched satellite tiles), the per-model build-duration budget, the retry/fallback logic on build failure, and the on-disk engine cache layout.
+
+**Out-of-research-scope (deferred to Plan-phase)**:
+- Operator-side CLI/desktop tool design (mission-prep tooling shape; CLI vs GUI; integration with QGC plan files / MAVProxy / Mission Planner equivalents).
+- Sector classification (active-conflict vs stable rear) heuristics + interface — used to decide AC-8.2 freshness threshold (6 mo vs 12 mo).
+- Tile age-stamping + freshness schema beyond what AC-8.2 + AC-NEW-6 already mandate.
+
+**Rationale for narrowing**:
+- The C6 and C7 closures already locked architectural recommendations (`periodic rebuild during pre-flight` and `build-on-deployed-Jetson at pre-flight`). What remains is mechanism confirmation, not candidate enumeration.
+- The deferred items are fixture/operator-tooling-class concerns. Their cross-coupling with the runtime architecture is mediated entirely by the descriptor-cache file and the TensorRT engine cache file — both fixed by the in-scope confirmations. Operator tool design can iterate freely at Plan-phase without touching runtime contracts.
+- Aligns with the C9-restructure precedent: keep research focused on architecture-binding decisions; push fixture/tooling decisions to the phases that own them.
+
+**Effective changes**:
+- **Component Areas table**: C10 row preserved with reduced scope. Per-FC details below.
+- **`Required outputs` for C10 in the table**: narrows from `Tooling (operator-side) to pull tiles from Suite Sat Service for an operational area, classify active-conflict vs stable rear, age-stamp, populate descriptor index` to `Confirmed orchestration mechanism for descriptor-cache rebuild + TensorRT engine build at pre-flight; on-disk artifact format(s); time/memory budget; failure-mode + retry behavior`.
+- **Cross-component gates**: D-C6-3 and D-C7-7 remain owned jointly with C10; new C10-internal decisions D-C10-x will be added at C10 batch 1 closure.
+- **SQ5 interleaving**: limited C10 SQ5 facts (failure modes during pre-flight build/rebuild) collected during this batch.
+
+**Carryforward to Plan-phase** — operator-tooling design issues preserved here so Plan-phase has a starting list:
+- Tool shape: integrate as a sub-command of Mission Planner / QGC plan-file workflow vs standalone CLI vs lightweight desktop GUI.
+- Sector-classification source: operator-marked geofence polygons vs Suite Sat Service metadata vs hybrid.
+- Tile age-stamping: per-tile capture date in manifest (already mandated by restrictions.md) vs additional sector-class tag vs full audit trail per AC-NEW-7.
+- Freshness pipeline: when to re-pull from Suite Sat Service (every flight, weekly, on operator demand, on sector-class change).
+
+## Next Step
+
+SQ1 ✓ → SQ2 ✓ (with three architectural decisions resolved) → **SQ3+SQ4 per component (C1→C8)** ✓ → **C10 batch 1 in progress (cross-coupling minimal scope, 2 sub-areas: D-C6-3 + D-C7-7 confirmation)** → SQ5 interleaved → SQ8 → SQ9 synthesis at engine Step 8.
+
+(SQ7 deferred to Test Spec per C9 restructure; C9 dropped; C10 operator-tooling-design deferred to Plan-phase per the C10 scope restructure above.)
+
+Pipeline shape (final, post-C10-restructure): `C1 (VIO) → C2 (VPR) → Top-N re-rank by inlier count → C3 (matcher) → AdHoP-conditional refinement → C4 (PnP+RANSAC+LM) → C5 (estimator) → C8 (FC adapter)` with C6 (cache, 2D ortho) + C7 (Jetson runtime) + C10 (pre-flight orchestration: descriptor-cache rebuild + TensorRT engine build) cross-cutting.
+
+First C1 (VIO) candidate batch: VINS-Mono / VINS-Fusion / OpenVINS / OKVIS2 / DROID-SLAM / DPVO / pure-VO baseline (RTAB-Map and ORB-SLAM3 already pruned by Fact #16). Per-mode `context7` capability verification mandatory for every lead library/SDK candidate.

_docs/00_research/01_source_registry/00_summary.md

@@ -0,0 +1,172 @@
+# Source Registry — Summary & Index
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation).
+> Critical-novelty sensitivity per Step 0.5 in `../00_question_decomposition.md`. Time windows applied:
+> - **Lead-candidate / SOTA claims**: prefer sources within last 6 months; up to 18 months if older is the official authority.
+> - **Library/SDK API behaviour**: must reflect the currently shipped version at search time (`context7` mandatory per lead candidate).
+> - **Established baselines** (KLT, RANSAC, EKF, ORB, SIFT, GTSAM): no time window.
+>
+> Investigation order saved in `../00_question_decomposition.md` → "Next Step": SQ6 → SQ1 → SQ2 → SQ3+SQ4 per component (C1→C8) ✓ → C10 next → SQ5 interleaved → SQ8 → SQ9 synthesis at engine Step 8. **SQ7 (datasets / SITL / replay) deferred to Test Spec (greenfield Step 5) per 2026-05-08 C9 / SQ7 restructure** — see `../00_question_decomposition.md` → "C9 / SQ7 Restructure" section.
+>
+> This folder replaces the previous monolithic `01_source_registry.md`. The full per-source description for any source `#N` in the table below lives in the category file linked in its row.
+
+## Category Index
+
+| Category | File | Sources | Status |
+|---|---|---|---|
+| SQ6 — ArduPilot Plane vs iNav external positioning | [`SQ6_external_positioning.md`](SQ6_external_positioning.md) | #1–#24 | Saturated for protocol-level architectural decision |
+| SQ1 — Existing GPS-denied UAV systems | [`SQ1_existing_systems.md`](SQ1_existing_systems.md) | #25–#37 | Saturated |
+| SQ2 — Canonical pipeline decomposition | [`SQ2_canonical_pipeline.md`](SQ2_canonical_pipeline.md) | #38–#42 | Saturated |
+| C1 — VIO candidates | [`C1_vio.md`](C1_vio.md) | #43–#56 | Closed at documentary level |
+| C2 — VPR candidates | [`C2_vpr.md`](C2_vpr.md) | #57–#68 | Mandatory pre-screen complete (5/5) |
+| C3 — Matcher candidates | [`C3_matchers.md`](C3_matchers.md) | #69–#81 | Closed at documentary level |
+| C4 — Pose estimation candidates | [`C4_pose_estimation.md`](C4_pose_estimation.md) | #82–#87 | Closed at 3/N |
+| C5 — State estimator / sensor fusion candidates | [`C5_state_estimator.md`](C5_state_estimator.md) | #88–#91 | Closed at 2/N (batch 1 closed) |
+| C6 — Tile cache + spatial index candidates | [`C6_tile_cache_spatial_index.md`](C6_tile_cache_spatial_index.md) | #92–#98 | Closed at 2/N (batch 1 closed) — Cand 1 (mirror-suite-pattern) RECOMMENDED PRIMARY; Cand 2 (PostGIS+pgvector) DEFERRED secondary |
+| C7 — On-Jetson inference runtime candidates | [`C7_inference_runtime.md`](C7_inference_runtime.md) | #99–#105 | Closed at 3/N (batch 1 closed 2026-05-08) — Cand 1 (TensorRT native) RECOMMENDED PRIMARY; Cand 2 (ONNX Runtime + TRT EP) modern-competitive-lead-cross-architecture-portability; Cand 3 (pure PyTorch FP16) mandatory simple-baseline |
+| C8 — MAVLink / MSP2 FC adapter candidates | [`C8_fc_adapter.md`](C8_fc_adapter.md) | #106–#113 | Closed at 3/N (batch 1 closed 2026-05-08) — Cand 1 (pymavlink → MAVLink GPS_INPUT) RECOMMENDED PRIMARY for ArduPilot Plane; Cand 2 (MSP2_SENSOR_GPS via Python MSP V2) RECOMMENDED PRIMARY for iNav (locked SQ6 + AC-4.3 transport); Cand 3 (UBX impersonation via pyubx2 NAV-PVT) DEFERRED secondary for iNav after comparative-improvement verdict |
+| C10 — Pre-flight cache provisioning (CROSS-COUPLING MINIMAL scope per 2026-05-08 user choice C; D-C6-3 + D-C7-7 confirmation pipelines only, operator tooling deferred to Plan-phase) | [`C10_preflight_provisioning.md`](C10_preflight_provisioning.md) | #114–#121 | Closed at 2/N (batch 1 closed 2026-05-08) — D-C6-3 confirmation: direct `faiss.write_index`/`faiss.read_index` Python API + `python-atomicwrites` + content-hash verification gate at takeoff (FAISS MIT, atomicwrites MIT); D-C7-7 confirmation: hybrid Polygraphy CLI primary + `trtexec` for cache-reuse fast rebuilds + direct `IBuilderConfig` Python API escape hatch (Polygraphy + TensorRT 10.x Apache-2.0 throughout) |
+| **Mode B addendum (2026-05-08)** — solution_draft01 assessment | [`MODEB_addendum.md`](MODEB_addendum.md) | **#122–#131** (10 sources) | New sources gathered for Mode B findings F1–F20: VINS-Mono GPL-3.0 LICENCE confirmation (#122), MegaLoc + UltraVPR + AirZoo aerial-VPR successor candidates (#123, #124, #125), CVE-2026-1579 MAVLink no-default-auth + CVE-2025-53644 OpenCV crafted-JPEG (#126, #127), ArduPilot MAVLink2 message-signing + iNav signing-gap (#128, #129), ArduPilot `MAV_CMD_SET_EKF_SOURCE_SET` no-deployed-GCS-implementer re-verification (#130), XoFTR + 2026 SAR-optical 24-matcher benchmark (#131). |
+
+## Investigation Status
+
+| Sub-question | Status | Notes |
+|---|---|---|
+| SQ6 — ArduPilot vs iNav external positioning | **Saturated for protocol-level architectural decision** (further detail deferred to SQ8 for spoofing-side fields and to design phase for SITL parameter tuning) | Major finding: iNav has no inbound external-positioning MAVLink handler; AC-4.3 wording must be revised. See `../02_fact_cards/SQ6_fc_external_positioning.md` "SQ6 Conclusions". |
+| SQ1 — Existing GPS-denied UAV systems | **Saturated.** 13 sources logged across academic / open-source / commercial / defense-program / Ukraine-practitioner. Closest peer system: Twist Robotics OSCAR (deployed in Ukraine). Closest open-source pipeline-match: snktshrma/ngps_flight (NGPS, ArduPilot GSoC 2024 — LightGlue+SuperPoint+UKF+VISION_POSITION_ESTIMATE). Closest deployed commercial: Auterion Artemis (Skynode N + Visual Navigation, Ukraine-tested, 1000-mile range). | See `../02_fact_cards/SQ1_existing_systems.md` cluster + working summary. |
+| SQ2 — Canonical pipeline decomposition | **Saturated.** 5 surveys/benchmarks logged (Skoltech aerial VPR, U.Maine cross-view, OrthoLoC 2.5D geodata, AnyVisLoc low-altitude multi-view, NUDT 2026 sciopen survey). All converge on **`retrieval → matching → pose-estimation`** hierarchical framework with VIO/IMU as auxiliary. Two new architectural facts added to C1–C10: (a) **AdHoP-style perspective-refinement loop** between matching and PnP (+63% translation accuracy, method-agnostic), (b) **DSM 2.5D dependency** for full 6-DoF on aerial-to-satellite (must be resolved with the Suite Sat Service or accepted as a 3-DoF degraded mode). Practitioner runtime evidence: AnyLoc on RTX 3090 = 0.63s/descriptor, SuperGlue re-rank = 17–25s; on Jetson Orin Nano these are non-viable for our 400 ms p95 budget — must restrict to lightweight VPR (e.g., MixVPR / SALAD class) + LightGlue/XFeat-class matchers. See `../02_fact_cards/SQ2_canonical_pipeline.md` "SQ2 Conclusions". |
+| SQ3+SQ4 — Per-component candidates (C1–C10) | **In progress** — C1 (VIO) **CLOSED** at documentary level (Sources #43–#56). C2 (VPR) — **mandatory pre-screen COMPLETE at documentary level (5 of 5 candidates)**: MixVPR (Sources #57+#58), SALAD (Sources #59+#60+#61), SelaVPR (Sources #62+#63), NetVLAD (Sources #64+#65+#66), **EigenPlaces (Sources #67+#68 — closure 2026-05-08)**. All five mandatory candidates have per-mode API capability verification ✅, per-numbered-Restriction × per-numbered-AC sub-matrix written, and `../06_component_fit_matrix/C2_vpr.md` rows populated. **Conditional pre-screen candidates (AnyLoc / BoQ / DINOv2-VLAD)** are GATED on a prerequisite **INT8 quantization survey** before they can be added to per-mode rows (per Fact #26 pre-screen rule). C3 closed at documentary level (Sources #69–#81). C4 closed at 3/N (Sources #82–#87). **C5 CLOSED at 2/N — batch 1 closed 2026-05-08** (mandatory simple-baseline = Manual ESKF Solà 2017 [Sources #88–#89]; modern-competitive-lead-factor-graph = GTSAM iSAM2 + ImuFactor + smart factors + Marginals [Sources #90–#91]). **C6 CLOSED at 2/N — batch 1 closed 2026-05-08** (Cand 1 RECOMMENDED PRIMARY = mirror-of-suite-satellite-provider pattern: PostgreSQL btree + bytea + FAISS HNSW + filesystem [Sources #92+#96+#97+#98]; Cand 2 DEFERRED secondary = PostGIS GiST + pgvector HNSW + filesystem [Sources #94+#95]; Source #93 = PostgreSQL btree multicolumn-indexes docs cross-cite). **C7 CLOSED at 3/N — batch 1 closed 2026-05-08** (Cand 1 RECOMMENDED PRIMARY = TensorRT native [Sources #99+#104+#105]; Cand 2 modern-competitive-lead-cross-architecture-portability = ONNX Runtime + TRT EP [Source #100 + #103]; Cand 3 mandatory simple-baseline = pure PyTorch FP16 [Source #101]; Source #102 = YOLO26 Jetson Orin Nano Super benchmark; Source #103 = LightGlue+TRT+FP8 quantization-sensitivity finding driving D-C7-6 cross-component precision policy). **C8 CLOSED at 3/N — batch 1 closed 2026-05-08** (Cand 1 RECOMMENDED PRIMARY for ArduPilot = pymavlink → MAVLink GPS_INPUT msg 232 cooperative-path [Sources #106+#107 + cross-cite SQ6 Source #4 AP_GPS_MAV.cpp ingestion-path]; Cand 2 RECOMMENDED PRIMARY for iNav = MSP2_SENSOR_GPS id 7939 / 0x1F03 via Python MSP V2 implementation [Sources #111+#112+#113 + cross-cite SQ6 Source #12+#13]; Cand 3 DEFERRED secondary for iNav = UBX impersonation via pyubx2 NAV-PVT [Sources #108+#109+#110 + cross-cite SQ6 Fact #10] with comparative-improvement verdict that does NOT clear user's "significant-improvement-only" bar over Cand 2; mid-batch correction via c8_inav_recovery=B preserved locked SQ6 + AC-4.3 + restrictions.md verdicts). **C9 DROPPED** from research scope per 2026-05-08 SQ7/C9 restructure (datasets/SITL/replay deferred to Test Spec greenfield Step 5). **C10 CLOSED at 2/N — batch 1 closed 2026-05-08** under CROSS-COUPLING MINIMAL scope per 2026-05-08 user choice C (operator CLI/desktop tooling, sector classification, freshness pipeline deferred to Plan-phase): D-C6-3 confirmation = direct `faiss.write_index`/`faiss.read_index` Python API + `python-atomicwrites` + content-hash (SHA-256) verification gate at takeoff load + `IO_FLAG_MMAP_IFC` mmap [Sources #114+#115+#116]; D-C7-7 confirmation = hybrid Polygraphy CLI primary for INT8-calibrating builds + `trtexec` for cache-reuse fast rebuilds + direct `IBuilderConfig` Python API escape hatch [Sources #117+#118+#119+#120+#121]; **no further C10 batches required at the research layer** — operator tooling design enters at Plan-phase. | See `../02_fact_cards/C1_vio.md` + `../02_fact_cards/C2_vpr.md` + `../02_fact_cards/C3_matchers.md` + `../02_fact_cards/C4_pose_estimation.md` + `../02_fact_cards/C5_state_estimator.md` + `../02_fact_cards/C6_tile_cache_spatial_index.md` + `../02_fact_cards/C7_inference_runtime.md` clusters; `../06_component_fit_matrix/C{1..7}_*.md` rows. |
+| SQ5 — Failure modes / deployment lessons | Not started (interleaved with SQ3/SQ4) | |
+| SQ7 — Datasets, SITL, replay environments | **Deferred to Test Spec (greenfield Step 5)** per 2026-05-08 C9 / SQ7 restructure | Fixture-class / test-infra-class — not researched in this Mode A run. Carryforward payload preserved in `../00_question_decomposition.md` → "C9 / SQ7 Restructure" section. |
+| SQ8 — Safety considerations (AC-NEW-4 / AC-NEW-7) | Not started | Carries the AP_GPS spoofing-signal probe deferred from SQ6. |
+| SQ9 — End-to-end synthesis | Step 8 of engine (deferred) | |
+
+---
+
+## Source Summary Table
+
+Compact one-line index across all 121 sources. For full per-source description, follow the **File** link.
+
+| # | Title | Tier | File |
+|---|---|---|---|
+| 1 | Non-GPS Navigation — Plane documentation | L1 | [SQ6](SQ6_external_positioning.md) |
+| 2 | GPS / Non-GPS Transitions — Plane documentation | L1 | [SQ6](SQ6_external_positioning.md) |
+| 3 | EKF Source Selection and Switching — Plane documentation | L1 | [SQ6](SQ6_external_positioning.md) |
+| 4 | ArduPilot AP_GPS_MAV.cpp (master) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 5 | ArduPilot PR #28750 — AP_NavEKF3 EK3_OPTION bits (GPS-denied testing) | L2 | [SQ6](SQ6_external_positioning.md) |
+| 6 | ArduPilot Issue #15859 — EKF3 source switching (GPS↔NonGPS) | L4 | [SQ6](SQ6_external_positioning.md) |
+| 7 | ArduPilot Issue #27193 — EK3 Source Switching wrong frame for GUIDED | L4 | [SQ6](SQ6_external_positioning.md) |
+| 8 | ArduPilot Issue #23485 — fuse only External Nav Velocities | L4 | [SQ6](SQ6_external_positioning.md) |
+| 9 | iNavFlight/inav telemetry/mavlink.c (master inbound switch) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 10 | iNav Wiki — MAVLink (frogmane edited 2025-12-11) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 11 | iNav Wiki — GPS and Compass setup | L1 | [SQ6](SQ6_external_positioning.md) |
+| 12 | iNavFlight/inav docs/development/msp/README.md (MSP message reference) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 13 | iNavFlight/inav src/main/io/gps.c + target/common.h (master) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 14 | iNav Issue #10141 — dual GPS support | L4 | [SQ6](SQ6_external_positioning.md) |
+| 15 | iNav docs/GPS_fix_estimation.md (master) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 16 | iNav docs/Settings.md (master) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 17 | iNav Issue #10588 — DeadReckoning weird behaviour during GPS outage | L4 | [SQ6](SQ6_external_positioning.md) |
+| 18 | iNav Release 8.0.0 (highlights, Dec 2024) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 19 | iNav Release 9.0.0 / 9.0.1 + Release Notes wiki | L1 | [SQ6](SQ6_external_positioning.md) |
+| 20 | MAVLink common message set — GPS_RAW_INT (24) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 21 | MAVLink PR #2110 — gps: add status and integrity information | L2 | [SQ6](SQ6_external_positioning.md) |
+| 22 | AirDroper — GNSS Spoofing Filter companion device | L3 | [SQ6](SQ6_external_positioning.md) |
+| 23 | ArduPilot PR #24135 — EKF3 robust to bad IMU and lane-switching | L2 | [SQ6](SQ6_external_positioning.md) |
+| 24 | ArduPilot AP_NavEKF3 — VehicleStatus.cpp + AP_NavEKF3.cpp (master) | L1 | [SQ6](SQ6_external_positioning.md) |
+| 25 | Twist Robotics OSCAR — visual navigation system (Ukraine deployment) | L2 | [SQ1](SQ1_existing_systems.md) |
+| 26 | Ukraine Drones with Vision-Based Navigation Past Heavy Jamming (TWZ) | L2 | [SQ1](SQ1_existing_systems.md) |
+| 27 | Ukraine's Ruta Missile Drone EW-Immune Navigation (Defense Express) | L2 | [SQ1](SQ1_existing_systems.md) |
+| 28 | Kilometer-Scale GNSS-Denied UAV Navigation via Heightmap Gradients | L1 | [SQ1](SQ1_existing_systems.md) |
+| 29 | Hierarchical Image Matching for UAV Absolute Visual Localization | L1 | [SQ1](SQ1_existing_systems.md) |
+| 30 | Raptor — GPS-Denied UAV Navigation & Coordinate Extraction (Vantor) | L2 | [SQ1](SQ1_existing_systems.md) |
+| 31 | Auterion Artemis program — long-range deep-strike completion | L1 | [SQ1](SQ1_existing_systems.md) |
+| 32 | Auterion Skynode N — AI/CV for small autonomous systems | L2 | [SQ1](SQ1_existing_systems.md) |
+| 33 | snktshrma/ngps_flight — NGPS for ArduPilot (GSoC 2024) | L1 | [SQ1](SQ1_existing_systems.md) |
+| 34 | AerialExtreMatch — benchmark for extreme-view image matching/localization | L1 | [SQ1](SQ1_existing_systems.md) |
+| 35 | DARPA Fast Lightweight Autonomy (FLA) program page + T&E review | L1 | [SQ1](SQ1_existing_systems.md) |
+| 36 | DSMAC / TERCOM lineage — DTIC ADA315439 | L1 | [SQ1](SQ1_existing_systems.md) |
+| 37 | Electronic Warfare in Ukraine — Ukraine War Analytics | L3 | [SQ1](SQ1_existing_systems.md) |
+| 38 | VPR for Aerial Imagery: A Survey (Skoltech, Moskalenko et al.) | L1 | [SQ2](SQ2_canonical_pipeline.md) |
+| 39 | Cross-View Geo-Localization: A Survey (U. Maine) | L1 | [SQ2](SQ2_canonical_pipeline.md) |
+| 40 | OrthoLoC: UAV 6-DoF Localization with Orthographic Geodata | L1 | [SQ2](SQ2_canonical_pipeline.md) |
+| 41 | AnyVisLoc — UAV visual localization, low-altitude multi-view | L1 | [SQ2](SQ2_canonical_pipeline.md) |
+| 42 | NUDT 2026 — survey on absolute visual localization for low-altitude UAV | L1 | [SQ2](SQ2_canonical_pipeline.md) |
+| 43 | VINS-Mono — robust monocular visual-inertial state estimator | L1 | [C1](C1_vio.md) |
+| 44 | VINS-Fusion — optimization-based multi-sensor state estimator | L1 | [C1](C1_vio.md) |
+| 45 | OpenVINS — open-source VI navigation research platform | L1 | [C1](C1_vio.md) |
+| 46 | Run VIO on NVIDIA Jetson — KAIST benchmark | L1 | [C1](C1_vio.md) |
+| 47 | OKVIS2 — realtime scalable VI-SLAM with loop closure | L1 | [C1](C1_vio.md) |
+| 48 | OKVIS2-X — open keyframe VI-SLAM with dense depth | L1 | [C1](C1_vio.md) |
+| 49 | Kimera-VIO — VIO with SLAM + 3D mesh (MIT-SPARK, BSD) | L1 | [C1](C1_vio.md) |
+| 50 | DROID-SLAM — deep visual SLAM (princeton-vl) | L1 | [C1](C1_vio.md) |
+| 51 | DPVO / DPV-SLAM — deep patch visual odometry | L1 | [C1](C1_vio.md) |
+| 52 | DPVO-QAT++ — heterogeneous QAT + CUDA kernel fusion for DPVO | L2 | [C1](C1_vio.md) |
+| 53 | Pure-VO baseline — KLT optical flow + 5-point/homography RANSAC (OpenCV) | L1 | [C1](C1_vio.md) |
+| 54 | OpenVINS — context7 per-mode capability lookup (`/rpng/open_vins`) | L1 | [C1](C1_vio.md) |
+| 55 | VINS-Mono README + VINS-Fusion context7 per-mode lookup | L1 | [C1](C1_vio.md) |
+| 56 | OKVIS2 — official README (`smartroboticslab/okvis2`, main) | L1 | [C1](C1_vio.md) |
+| 57 | OpenVPRLab — open-source VPR framework (MixVPR / BoQ / NetVLAD / GeM) | L1 | [C2](C2_vpr.md) |
+| 58 | MixVPR canonical paper (WACV 2023, arXiv:2303.02190) | L1 | [C2](C2_vpr.md) |
+| 59 | SALAD canonical implementation (`serizba/salad`, GPL-3.0) | L1 | [C2](C2_vpr.md) |
+| 60 | SALAD canonical paper — Optimal Transport Aggregation (CVPR 2024) | L1 | [C2](C2_vpr.md) |
+| 61 | OpenVPRLab DinoV2 backbone — context7 cross-source for ViT-B/14 | L1 | [C2](C2_vpr.md) |
+| 62 | SelaVPR canonical implementation (`Lu-Feng/SelaVPR`, MIT) | L1 | [C2](C2_vpr.md) |
+| 63 | SelaVPR canonical paper (ICLR 2024, arXiv:2402.14505) | L1 | [C2](C2_vpr.md) |
+| 64 | NetVLAD canonical implementation `Relja/netvlad` v1.03 (MIT) | L1 | [C2](C2_vpr.md) |
+| 65 | NetVLAD modern PyTorch reproduction `Nanne/pytorch-NetVlad` | L2 | [C2](C2_vpr.md) |
+| 66 | NetVLAD canonical paper (CVPR 2016 / TPAMI 2018, arXiv:1511.07247) | L1 | [C2](C2_vpr.md) |
+| 67 | EigenPlaces canonical implementation (`gmberton/EigenPlaces`, MIT) | L1 | [C2](C2_vpr.md) |
+| 68 | EigenPlaces canonical paper (ICCV 2023, arXiv:2308.10832) | L1 | [C2](C2_vpr.md) |
+| 69 | LightGlue — context7 per-mode capability lookup (`/cvg/lightglue`) | L1 | [C3](C3_matchers.md) |
+| 70 | LightGlue canonical implementation (`cvg/LightGlue`) | L1 | [C3](C3_matchers.md) |
+| 71 | LightGlue canonical paper (ICCV 2023, arXiv:2306.13643) | L1 | [C3](C3_matchers.md) |
+| 72 | LightGlue HuggingFace Transformers integration | L1 | [C3](C3_matchers.md) |
+| 73 | LightGlue-ONNX — `fabio-sim/LightGlue-ONNX` (Jetson TensorRT path) | L2 | [C3](C3_matchers.md) |
+| 74 | ALIKED canonical implementation (`Shiaoming/ALIKED`) | L1 | [C3](C3_matchers.md) |
+| 75 | ALIKED canonical paper (TIM 2023, arXiv:2304.03608) | L1 | [C3](C3_matchers.md) |
+| 76 | DISK canonical implementation (`cvlab-epfl/disk`, Apache-2.0) | L1 | [C3](C3_matchers.md) |
+| 77 | DISK canonical paper — RL-trained local features (NeurIPS 2020) | L1 | [C3](C3_matchers.md) |
+| 78 | SuperGlue canonical implementation (`magicleap/SuperGluePretrainedNetwork`) | L1 | [C3](C3_matchers.md) |
+| 79 | SuperGlue canonical paper — graph-NN feature matching (CVPR 2020) | L1 | [C3](C3_matchers.md) |
+| 80 | XFeat canonical implementation (`verlab/accelerated_features`, Apache-2.0) | L1 | [C3](C3_matchers.md) |
+| 81 | XFeat canonical paper — accelerated features (CVPR 2024) | L1 | [C3](C3_matchers.md) |
+| 82 | OpenCV canonical implementation — `opencv/opencv` (calib3d module) | L1 | [C4](C4_pose_estimation.md) |
+| 83 | OpenCV 4.x calib3d module canonical documentation | L1 | [C4](C4_pose_estimation.md) |
+| 84 | OpenGV canonical implementation (`laurentkneip/opengv`) | L1 | [C4](C4_pose_estimation.md) |
+| 85 | OpenGV canonical Doxygen documentation portal | L1 | [C4](C4_pose_estimation.md) |
+| 86 | GTSAM canonical implementation (`borglab/gtsam`, BSD-3) | L1 | [C4](C4_pose_estimation.md) |
+| 87 | GTSAM canonical Python documentation via context7 | L1 | [C4](C4_pose_estimation.md) |
+| 88 | Solà 2017 — "Quaternion kinematics for the error-state Kalman filter" (arXiv:1711.02508) | L1 | [C5](C5_state_estimator.md) |
+| 89 | Reference open-source ESKF implementations (canonical-paper-derived) | L2 | [C5](C5_state_estimator.md) |
+| 90 | GTSAM `ImuFactor` / `CombinedImuFactor` / `PreintegratedImuMeasurements` / `PreintegratedCombinedMeasurements` (context7 indexed) | L1 | [C5](C5_state_estimator.md) |
+| 91 | GTSAM `ISAM2` / `IncrementalFixedLagSmoother` / `Marginals` with iSAM2 results (context7 indexed) | L1 | [C5](C5_state_estimator.md) |
+| 92 | Parent-suite `satellite-provider` existing pattern (PostgreSQL + Dapper + filesystem tile storage; verified directly) | L1 | [C6](C6_tile_cache_spatial_index.md) |
+| 93 | PostgreSQL 16 official documentation — Multicolumn Indexes + btree access method | L1 | [C6](C6_tile_cache_spatial_index.md) |
+| 94 | PostGIS official documentation — GiST + KNN distance ordering + ST_DWithin | L1 | [C6](C6_tile_cache_spatial_index.md) |
+| 95 | pgvector official documentation — HNSW index API (context7 + canonical README) | L1 | [C6](C6_tile_cache_spatial_index.md) |
+| 96 | FAISS official documentation — IndexFlatL2 / IndexHNSWFlat / IndexIVFFlat (context7 indexed) | L1 | [C6](C6_tile_cache_spatial_index.md) |
+| 97 | Postgres on NVIDIA Jetson Orin Nano — March 2026 Medium article + Coding Steve minimal-config guide | L2 | [C6](C6_tile_cache_spatial_index.md) |
+| 98 | Slippy Map Tilenames — OpenStreetMap canonical specification (Web Mercator XYZ) | L1 | [C6](C6_tile_cache_spatial_index.md) |
+| 99 | NVIDIA TensorRT 10.x official documentation portal (context7-indexed `/nvidia/tensorrt`) | L1 | [C7](C7_inference_runtime.md) |
+| 100 | Microsoft ONNX Runtime official documentation (context7-indexed `/microsoft/onnxruntime`) + Jetson AI Lab community wheel index | L1 | [C7](C7_inference_runtime.md) |
+| 101 | PyTorch official documentation (context7-indexed `/pytorch/pytorch`) + Jetson AI Lab PyTorch wheel availability for JetPack 6 | L1 | [C7](C7_inference_runtime.md) |
+| 102 | Ultralytics YOLO26 benchmark suite on Jetson Orin Nano Super (April 2026) | L2 | [C7](C7_inference_runtime.md) |
+| 103 | LightGlue ONNX Runtime + TensorRT acceleration + FP8 ModelOpt quantization findings (Fabio Sim's Journal) | L2 | [C7](C7_inference_runtime.md) |
+| 104 | JetPack SDK release notes (NVIDIA official) — JetPack 6.0 / 6.1 / 6.2 version matrix | L1 | [C7](C7_inference_runtime.md) |
+| 105 | TensorRT-on-Jetson canonical install constraints (Ultralytics issue reports + NVIDIA forum) | L2 | [C7](C7_inference_runtime.md) |
+| 106 | ArduPilot Pymavlink (context7-indexed `/ardupilot/pymavlink`) — canonical Python MAVLink stack | L1 | [C8](C8_fc_adapter.md) |
+| 107 | ArduPilot Plane Non-GPS Position Estimation + MAVProxy GPS Input module dev docs (`GPS1_TYPE=14`, `EK3_SRC1_POSXY=3`) | L1 | [C8](C8_fc_adapter.md) |
+| 108 | pyubx2 (context7-indexed `/semuconsulting/pyubx2`) — canonical Python UBX/NMEA/RTCM3 parser | L1 | [C8](C8_fc_adapter.md) |
+| 109 | u-blox NEO-M9N Integration Manual (UBX-19014286) + u-blox 8/M8 Receiver Description (UBX-13003221) — UBX-NAV-PVT canonical specification | L1 | [C8](C8_fc_adapter.md) |
+| 110 | iNav `gps_ublox.c` source (master) — UBX validation gates `gpsMapFixType()` requires `flags & 0x01 = 1` AND `fixType ∈ {2,3}` | L1 | [C8](C8_fc_adapter.md) |
+| 111 | iNav `docs/development/msp/README.md` (master) — `MSP2_SENSOR_GPS (7939 / 0x1F03)` canonical 36-byte payload spec | L1 | [C8](C8_fc_adapter.md) |
+| 112 | Python MSP2 implementations: YAMSPy + INAV-Toolkit `inav_msp.py` (MSP V2 `msp_v2_encode` with CRC-8 DVB-S2) | L2 | [C8](C8_fc_adapter.md) |
+| 113 | iNav `src/main/msp/msp_protocol_v2_sensor.h` (master) — MSP V2 sensor command-ID range (0x1F00-0x1FFF) | L1 | [C8](C8_fc_adapter.md) |
+| 114 | FAISS `write_index` / `read_index` Python API + on-disk format + security warning (canonical wiki + context7) | L1 | [C10](C10_preflight_provisioning.md) |
+| 115 | FAISS IndexHNSWFlat per-vector memory + on-disk file size formula (Discussions #3953 + C++ API docs) | L2 | [C10](C10_preflight_provisioning.md) |
+| 116 | Python atomic file write pattern (gocept blog + python-atomicwrites docs + Python Issue 8604) | L2 | [C10](C10_preflight_provisioning.md) |
+| 117 | Polygraphy `polygraphy convert` CLI for TensorRT INT8 engine build with calibration cache reuse (NVIDIA TensorRT repo + context7) | L1 | [C10](C10_preflight_provisioning.md) |
+| 118 | Polygraphy `Calibrator` class API — algo defaults + dynamic-shapes calibration profile + warning behavior (NVIDIA TRT/Polygraphy SDK docs) | L1 | [C10](C10_preflight_provisioning.md) |
+| 119 | `trtexec` CLI for one-off engine builds — INT8/FP16 flags + calibration cache support (NVIDIA TRT SDK docs) | L1 | [C10](C10_preflight_provisioning.md) |
+| 120 | TensorRT INT8 calibration corpus size guidance (~500-1000 images) — Jetson AGX Orin (vendor engineering guide) | L2 | [C10](C10_preflight_provisioning.md) |
+| 121 | Direct TensorRT `IBuilderConfig` + `IInt8EntropyCalibrator2` Python API (NVIDIA TRT Python API docs, cross-cite from C7 #105) | L1 | [C10](C10_preflight_provisioning.md) |

_docs/00_research/01_source_registry/C10_preflight_provisioning.md

@@ -0,0 +1,119 @@
+# Source Registry — C10: Pre-flight cache provisioning (cross-coupling minimal scope)
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation). Sources for C10 batch 1 (cross-coupling minimal: D-C6-3 descriptor-cache rebuild trigger pipeline + D-C7-7 TensorRT engine-build pipeline). Sibling registries: [SQ1](SQ1_existing_systems.md), [SQ2](SQ2_canonical_pipeline.md), [SQ6](SQ6_external_positioning.md), [C1](C1_vio.md), [C2](C2_vpr.md), [C3](C3_matchers.md), [C4](C4_pose_estimation.md), [C5](C5_state_estimator.md), [C6](C6_tile_cache_spatial_index.md), [C7](C7_inference_runtime.md), [C8](C8_fc_adapter.md). Index: [`00_summary.md`](00_summary.md).
+>
+> Source-tier definitions per `references/source-tiering.md`: L1 = official primary docs / source code / canonical specs; L2 = official blog posts, vendor SDK docs, peer-reviewed papers; L3 = community Q&A, tutorial sites, secondary commentary; L4 = forum posts, mailing-list threads, single-author blog posts.
+
+---
+
+## Source #114 — FAISS `write_index` / `read_index` Python API + on-disk format + security warning (L1 official)
+
+**URL**: <https://github.com/facebookresearch/faiss/wiki/Index-IO,-cloning-and-hyper-parameter-tuning> + context7 indexed at `/facebookresearch/faiss` (Benchmark Score consistent with C6 batch 1 Source #96 lookup)
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L1** — canonical FAISS GitHub Wiki + canonical context7-indexed documentation
+
+**Relevance**: Confirms `faiss.write_index(index, path)` + `faiss.read_index(path)` Python API for serializing IndexHNSWFlat to disk and loading it back; confirms `IO_FLAG_MMAP_IFC` enables memory-mapped loading for HNSW + IndexFlatCodes-derived classes (zero-copy load — important for the project's <5 s takeoff load budget); documents the explicit security warning "No attempt is made to check the correctness of loaded data. A faulty or malicious file could lead to out-of-memory errors or code execution. Users are responsible for verifying that files loaded with `read_index` have not been altered since being written by `write_index`." This warning binds directly to AC-NEW-7 (cache-poisoning safety) and motivates the project-side content-hash verification gate before takeoff load. Confirms FAISS C++ signature: `void write_index(Index* index, const char* filename)` / `Index* read_index(const char* filename)`.
+
+**Evidence quality**: ✅ High — L1 canonical FAISS docs. Direct API verification.
+
+---
+
+## Source #115 — FAISS IndexHNSWFlat per-vector memory + on-disk file size formula (L2 community + L1 cross-cite)
+
+**URL**: <https://github.com/facebookresearch/faiss/discussions/3953> + cross-cite <https://faiss.ai/cpp_api/struct/structfaiss_1_1IndexHNSWFlat.html>
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L2** — FAISS GitHub Discussions thread (maintainer-confirmed answer) + L1 canonical FAISS C++ API docs cross-cite
+
+**Relevance**: Confirms IndexHNSWFlat per-vector on-disk + RAM cost formula: `(vector_dim × 4 bytes) + (M × 4 bytes × 2) + overhead from graph layers and geometric reallocation`. For project's pinned VPR descriptor candidates (per D-C2-9 / D-C2-10 / D-C2-6 / D-C6-1 = halfvec): at 2048-D float32 + M=32 → 8192 + 256 = **8448 bytes/vector** (~845 MB on disk for 100K tiles); at 2048-D halfvec (2-byte storage per descriptor element) → 4096 + 256 = **4352 bytes/vector** (~430 MB on disk for 100K tiles); at 512-D halfvec + M=32 → 1024 + 256 = **1280 bytes/vector** (~130 MB on disk for 100K tiles); at 256-D halfvec + M=32 → 512 + 256 = **768 bytes/vector** (~80 MB on disk for 100K tiles). All variants well within AC-8.3 10 GB cache budget (assuming D-C2-10 EigenPlaces 512-D path or D-C6-1 halfvec mitigation). Supplementary cross-cite to C6 Fact #92 evidence base. **Load latency**: Issue #622 confirms post-load search performance is "slightly slower initially due to memory layout and cache effects" but identical results — implies a warmup-search-pass at takeoff after `read_index` would smooth p99 latency; aligns with the <5 s takeoff load budget (pure file read at ~430 MB / SATA SSD ~500 MB/s = <1 s; mmap path eliminates the read entirely).
+
+**Evidence quality**: ✅ High — formula matches FAISS source code in `IndexHNSW.cpp`; multiple maintainer-confirmed reproductions; conservative for project's pinned descriptor dimensions per D-C2-9/10/6 closures.
+
+---
+
+## Source #116 — Python atomic file write pattern: write-to-temp + fsync + atomic rename (L2 reference + L1 POSIX standard cross-cite)
+
+**URL**: <https://blog.gocept.com/2013/07/15/reliable-file-updates-with-python/> + <https://python-atomicwrites.readthedocs.io/en/stable> + Python tracker Issue 8604 <https://bugs.python.org/issue8604>
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L2** — well-known engineering blog reference + canonical Python package docs + Python core developer issue tracker
+
+**Relevance**: Documents the canonical Python crash-safe atomic file write pattern required for the project's pre-flight FAISS index file write (and TensorRT engine file write). The pattern is: (1) write to a temporary file in the same directory as target (ensures same filesystem so `os.rename` is atomic), (2) call `fsync(temp_fd)` to flush content + metadata to disk, (3) atomically rename via `os.rename(temp_path, target_path)`, (4) call `fsync` on the parent directory to flush the filename change to disk. Without this pattern, a power loss or process kill mid-write leaves a truncated/partial file that `faiss.read_index` will load successfully (no internal integrity check per Source #114 warning) and produce silently-wrong descriptor matches at takeoff — direct violation of AC-NEW-7 (cache-poisoning safety) + AC-3.3 (re-localization stability). The `python-atomicwrites` package provides this pattern with a simple API: `with atomic_write(path, overwrite=True) as f: ...`; pure-Python; trivially auditable; cross-platform (Windows + POSIX + macOS). On macOS specifically, must use `fcntl.fcntl(fd, fcntl.F_FULLFSYNC)` instead of `os.fsync()` to handle Apple's user-space write buffers — not relevant for the Jetson deployment target (Linux/JetPack). Project-side wrapper around `faiss.write_index` should use this pattern to safely write the FAISS cache file alongside content-hash verification.
+
+**Evidence quality**: ✅ High — pattern matches POSIX `rename(2)` atomicity guarantee; extensively documented; multiple production Python packages (atomicwrites, ruamel-yaml, etc.) implement it.
+
+---
+
+## Source #117 — Polygraphy `polygraphy convert` CLI for TensorRT INT8 engine build with calibration cache reuse (L1 official)
+
+**URL**: <https://github.com/NVIDIA/TensorRT/blob/main/tools/Polygraphy/examples/cli/convert/01_int8_calibration_in_tensorrt/README.md> + context7 indexed at `/websites/nvidia_deeplearning_tensorrt_static_polygraphy` (1041 code snippets, Benchmark Score 67.2, Source Reputation High)
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L1** — official NVIDIA TensorRT source repository documentation + canonical Polygraphy docs
+
+**Relevance**: Confirms Polygraphy as the canonical NVIDIA-blessed orchestration wrapper around TensorRT's engine build pipeline. Documents the canonical INT8 calibration workflow: first build with `--data-loader-script ./data_loader.py --calibration-cache identity_calib.cache` (computes scales + writes cache); subsequent builds with `--calibration-cache identity_calib.cache` (skips calibration step entirely — cache contains scales). Confirms Polygraphy's `Calibrator` class API: `data_loader` parameter (generator/iterable yielding `{input_name: numpy.ndarray}` dicts), `cache` parameter (calibration cache file path), `BaseClass` parameter (defaults to `trt.IInt8EntropyCalibrator2` — matches project's D-C7-2 + D-C7-6 lock), `algo` parameter (defaults to `trt.CalibrationAlgoType.MINMAX_CALIBRATION`). CLI supports `--int8 --fp16` mixed precision flags directly per project's D-C7-2 = (b) per-family precision policy. The full CLI invocation pattern for project: `polygraphy convert <model>.onnx --int8 --fp16 --data-loader-script ./calib_data_loader.py --calibration-cache <model>_calib.cache -o <model>_sm87_jp62_trt103_int8fp16.engine`. Polygraphy is bundled inside the TensorRT distribution (no separate install on Jetson — `pip install nvidia-pyindex && pip install polygraphy` or via TensorRT installer). Production-mature and cross-referenced from canonical TensorRT documentation.
+
+**Evidence quality**: ✅ High — official NVIDIA repository docs, multi-snippet context7 coverage, production-mature tooling.
+
+---
+
+## Source #118 — Polygraphy `Calibrator` class API — algo defaults + dynamic-shapes calibration profile + warning behavior (L1 official)
+
+**URL**: <https://docs.nvidia.com/deeplearning/tensorrt/latest/_static/polygraphy/backend/trt/calibrator.html> + <https://docs.nvidia.com/deeplearning/tensorrt/latest/_static/polygraphy/backend/trt/config.html>
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L1** — canonical NVIDIA TensorRT/Polygraphy SDK documentation
+
+**Relevance**: Confirms `Calibrator(data_loader, cache=None, BaseClass=IInt8EntropyCalibrator2, algo=CalibrationAlgoType.MINMAX_CALIBRATION, batch_size=None, quantile=None, regression_cutoff=None)` full signature. Documents two algorithm choices: `IInt8EntropyCalibrator2` (entropy-based; project D-C7-2 default; Polygraphy default) vs `IInt8MinMaxCalibrator` (min-max scaling). Documents dynamic-shapes behavior: "if calibration is run and the model has dynamic shapes, the last optimization profile will be used as the calibration profile" — relevant for project's matchers if any of them export with dynamic input shapes (D-C3-2 LightGlue ONNX export pathway). Documents `--data-loader-script` / `--data-loader-func-name` CLI flags for supplying custom calibration data. Documents the "Int8 Calibration is using randomly generated input data" warning that fires when `--int8` is set but neither `--data-loader-script` nor an existing `--calibration-cache` is supplied — operationalizes the D-C7-1 closure (real UAV nadir flight footage corpus) as a pre-flight build prerequisite. CLI also supports `--load-tactics` / `--save-tactics` for replaying tactic-search results across multiple builds (faster than re-running tactic profiling each build) — useful for the reference-Jetson-prebuilt-engine fallback path per D-C7-7.
+
+**Evidence quality**: ✅ High — canonical NVIDIA documentation, directly cited from polygraphy/tools/args/backend/trt/config source code.
+
+---
+
+## Source #119 — `trtexec` CLI for one-off engine builds — INT8/FP16 flags + calibration cache support (L1 official)
+
+**URL**: <https://docs.nvidia.com/deeplearning/tensorrt/latest/getting-started/quick-start-guide.html> + <https://docs.nvidia.com/deeplearning/tensorrt/latest/reference/command-line-programs.html>
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L1** — canonical NVIDIA TensorRT SDK documentation
+
+**Relevance**: Confirms `trtexec` as the simpler-but-less-flexible TensorRT engine build CLI bundled with every TensorRT installation. Canonical invocation: `trtexec --onnx=model.onnx --saveEngine=model.engine --fp16 --int8 --calib=calibration.cache --shapes=input:1x3x224x224`. Supports `--int8 --fp16` mixed precision (matches project's D-C7-2). Supports `--calib=<cache_path>` for INT8 calibration cache reuse (cache file format identical to Polygraphy's; the two tools are interoperable on the calibration cache layer). **Critical limitation vs Polygraphy**: `trtexec --int8` without `--calib` causes TRT to use random data for calibration (per TRT docs warning) — this collapses INT8 accuracy by ~5-15%. **Strength**: single-binary; no Python imports; no calibration data loader script required; perfect for emergency rebuilds when an existing calibration cache is available; perfect for ad-hoc benchmarking via `--iterations=N --useCudaGraph --noDataTransfers`. **Recommended role for project**: fallback orchestration tool when Polygraphy is unavailable OR when calibration cache is already shipped from a reference build (e.g., the prebuilt-engine fallback per D-C7-7).
+
+**Evidence quality**: ✅ High — canonical NVIDIA documentation; trtexec is bundled with TensorRT distributions and has been the canonical TensorRT CLI since TensorRT 5.x.
+
+---
+
+## Source #120 — TensorRT INT8 INT8 calibration corpus size guidance (~500-1000 images) — Jetson AGX Orin specific (L2 vendor)
+
+**URL**: <https://nvnexus.com/tensorrt-jetson-agx-orin-optimization-guide/>
+
+**Date accessed**: 2026-05-08
+
+**Tier**: **L2** — vendor-aligned engineering guide (TensorRT-on-Jetson specialist content), cross-cited from official NVIDIA Developer Forum patterns
+
+**Relevance**: Independent confirmation of the project's D-C7-1 closure: "INT8 optimization can double inference throughput on Jetson AGX Orin with minimal accuracy loss; calibration on representative input data (500-1000 images recommended)". Aligns with project's pinned 500-1500 sample range from C7 batch 1 Fact #94. Cross-cite to AGX Orin (server-class Jetson) — the project's deployment target is Orin Nano Super (smaller class), but the calibration-corpus-size guidance is governed by the model + INT8 entropy-statistics requirement, not by the Jetson SKU. **Conservative confirmation**: project's calibration corpus target of 500-1500 samples per D-C7-1 closure is sufficient by community-confirmed benchmarks.
+
+**Evidence quality**: ⚠️ Medium-High — L2 vendor-aligned source; aligns with multiple independent confirmations including NVIDIA Developer Forum threads and the canonical TensorRT INT8 calibration documentation; project's D-C7-1 closure already pinned this range from L1 sources.
+
+---
+
+## Source #121 — Direct TensorRT `IBuilderConfig` + `IInt8EntropyCalibrator2` Python API (L1 official, cross-cite from C7 Source #105)
+
+**URL**: <https://docs.nvidia.com/deeplearning/tensorrt/latest/_static/python/api/infer/Core/BuilderConfig.html> (cross-cite from C7 batch 1 Source #105 + Source #102)
+
+**Date accessed**: 2026-05-08 (cross-cite)
+
+**Tier**: **L1** — canonical NVIDIA TensorRT Python API documentation
+
+**Relevance**: Already cited in C7 batch 1 Source #102 + Source #105 (mode pinning for D-C7-2). Re-cited here for the C10 D-C7-7 confirmation context: confirms direct `IBuilderConfig` + `IInt8EntropyCalibrator2` Python API as the most-flexible-but-most-engineering-cost orchestration option. Pattern: instantiate `trt.Builder(logger)` → `builder.create_network(...)` → parse ONNX via `trt.OnnxParser` → instantiate `builder.create_builder_config()` → `config.set_flag(trt.BuilderFlag.INT8)` + `config.set_flag(trt.BuilderFlag.FP16)` → assign custom `Int8EntropyCalibrator2` subclass instance to `config.int8_calibrator` → `config.max_workspace_size = 1 << 30` (1 GB per D-C7-8) → `serialized_engine = builder.build_serialized_network(network, config)` → `with open(path, 'wb') as f: f.write(serialized_engine)`. **Used in C10 only as the per-model fallback path for the reference-Jetson-prebuilt-engine generation** (D-C7-7 fallback) when Polygraphy's data-loader-script abstraction is too rigid for an unusual model (e.g., LightGlue with dynamic-shape inputs requiring a custom calibration profile).
+
+**Evidence quality**: ✅ High — canonical NVIDIA Python API; cross-cite from existing C7 Source #105 reduces redundancy.
+
+---

_docs/00_research/01_source_registry/C1_vio.md

@@ -0,0 +1,192 @@
+# Source Registry — C1 — Visual / Visual-Inertial Odometry candidates
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation).
+> Critical-novelty sensitivity per Step 0.5 in `../00_question_decomposition.md`. Time windows applied:
+> - **Lead-candidate / SOTA claims**: prefer sources within last 6 months; up to 18 months if older is the official authority.
+> - **Library/SDK API behaviour**: must reflect the currently shipped version at search time (`context7` mandatory per lead candidate).
+> - **Established baselines** (KLT, RANSAC, EKF, ORB, SIFT, GTSAM): no time window.
+>
+> This file replaces a section of the previous monolithic `01_source_registry.md`. See `00_summary.md` for the full category index. Investigation order is tracked in `../00_question_decomposition.md` and the cross-category Investigation Status table in `00_summary.md`.
+
+---
+
+### Source #43
+- **Title**: VINS-Mono — A Robust and Versatile Monocular Visual-Inertial State Estimator (HKUST-Aerial-Robotics)
+- **Link**: https://github.com/HKUST-Aerial-Robotics/VINS-Mono ; LICENCE: https://github.com/HKUST-Aerial-Robotics/VINS-Mono/blob/master/LICENCE
+- **Tier**: L1 (canonical reference implementation; published in IEEE T-RO 2018 by Qin, Li, Shen)
+- **Publication Date**: original 2018; repository last meaningful update 2024-02-25 (per GitHub commit log; 2024-05-23 simulation-data commit only)
+- **Timeliness Status**: ⚠️ **Borderline.** ~24 months since the last meaningful master-branch commit at access time (2026-05-07). Established baseline that does NOT trigger Step 0.5's 18-month timeliness rejection because (a) IEEE T-RO publication is the canonical authority for the algorithm, (b) downstream forks (vins-mono-android, embedded variants) keep the algorithm class actively deployed.
+- **Version Info**: No GitHub releases / tags (master-branch-only project). Stars 5,829.
+- **Target Audience**: Mono+IMU VIO implementers; UAV state estimation researchers
+- **Research Boundary Match**: **Full match for the candidate's pinned mode** — monocular camera + IMU producing 6-DoF metric pose. The VINS-Mono README explicitly names this configuration as primary.
+- **Summary**: Optimization-based sliding-window monocular VIO. Features: efficient IMU pre-integration (Forster et al. 2017), automatic initialization, online camera-IMU extrinsic calibration, online camera-IMU temporal calibration, failure detection + recovery, loop detection (DBoW2-based), global pose graph optimization. Output is metric-scale 6-DoF pose at IMU rate (typically 100–200 Hz) with covariance from the optimization Hessian. **License: GPL-3.0 (copyleft viral)** — every binary distribution requires source disclosure for the entire linked binary; relevant for dual-use deployment if the companion image is sold or transferred to a customer.
+- **Related Sub-question**: SQ3+SQ4 / C1 lead candidate
+
+
+### Source #44
+- **Title**: VINS-Fusion — Optimization-based multi-sensor state estimator (HKUST-Aerial-Robotics)
+- **Link**: https://github.com/HKUST-Aerial-Robotics/VINS-Fusion ; LICENCE: https://github.com/HKUST-Aerial-Robotics/VINS-Fusion/blob/master/LICENCE
+- **Tier**: L1 (canonical reference; superset of VINS-Mono)
+- **Publication Date**: original 2019 (Qin, Cao, Pan, Shen — ICRA workshop / IROS); repository last update 2024-05-23
+- **Timeliness Status**: ⚠️ **Borderline.** ~24 months since the last update at access time. Same Step-0.5 reasoning as VINS-Mono — established class.
+- **Version Info**: master-branch-only. Stars 4,476. Top-ranked open-source stereo-VIO on KITTI Odometry as of January 2019.
+- **Target Audience**: Multi-sensor VIO implementers (mono+IMU, stereo, stereo+IMU, +GPS fusion)
+- **Research Boundary Match**: **Full match** for monocular+IMU mode. VINS-Fusion README explicitly enumerates four sensor configurations (mono+IMU, stereo, stereo+IMU, +GPS toy example).
+- **Summary**: Superset of VINS-Mono adding stereo and GPS-fusion modes. Same algorithmic core (sliding-window optimization with IMU pre-integration). Online spatial + temporal camera-IMU calibration; visual loop closure; ROS Kinetic/Melodic build dependency. **License: GPL-3.0** — same dual-use distribution constraint as VINS-Mono. Independent KAIST benchmark (Source #46) found VINS-Fusion CPU mode + VINS-Fusion-imu **fail to run** on Jetson TX2 (insufficient memory and CPU); GPU-accelerated VINS-Fusion-gpu does run on TX2. Implication for project: VINS-Fusion-imu on Jetson Orin Nano Super is feasible but not certain; needs MVE.
+- **Related Sub-question**: SQ3+SQ4 / C1 lead candidate
+
+
+### Source #45
+- **Title**: OpenVINS — An open source platform for visual-inertial navigation research (Robot Perception and Navigation Group, U. of Delaware — rpng)
+- **Link**: https://github.com/rpng/open_vins ; docs: https://docs.openvins.com/ ; LICENSE: https://github.com/rpng/open_vins/blob/master/LICENSE
+- **Tier**: L1 (canonical research implementation; ICRA 2020 paper Geneva, Eckenhoff, Lee, Yang, Huang)
+- **Publication Date**: original 2020; latest tagged release v2.7 = 2023-06; ongoing master-branch commits through 2024–2025 (latest issue threads through Feb 2025)
+- **Timeliness Status**: ✅ Currently valid (master branch active; latest tagged release ~35 months but library is in stable/maintenance mode with continued issue triage).
+- **Version Info**: Stars 2,828; 30 contributors; 12 releases. v2.7 is the current tagged stable.
+- **Target Audience**: MSCKF/EKF VIO implementers; researchers needing a reference MSCKF
+- **Research Boundary Match**: **Full match** for monocular+IMU mode. OpenVINS supports mono, stereo, multi-camera (1–N cameras) + IMU; mono is a documented first-class mode.
+- **Summary**: Modular MSCKF (Multi-State Constraint Kalman Filter) implementation built around an Extended Kalman filter that fuses inertial state with sparse visual feature tracks via the sliding-window MSCKF formulation (Mourikis & Roumeliotis 2007). Supports SLAM features (in-state landmarks) plus pure MSCKF features (out-of-state). ROS1 + ROS2 (Humble) builds documented; Jetson Orin Nano Dev Kit + JetPack 6 + ROS 2 Humble compilation **confirmed working** by community contributors (rpng/open_vins issue #421, fdcl-gwu/openvins_jetson_realsense Nov 2025 setup guide). **License: GPL-3.0** — same dual-use distribution constraint. Reported latency ~270 ms on Xavier NX (4-core, ARM, 40% CPU usage) per issue #164; needs Jetson-Orin-Nano-Super MVE for production budget verification.
+- **Related Sub-question**: SQ3+SQ4 / C1 lead candidate
+
+
+### Source #46
+- **Title**: Run Your Visual-Inertial Odometry on NVIDIA Jetson — Benchmark Tests on a Micro Aerial Vehicle (Jeon, Jung, Lee, Choi, Myung — KAIST)
+- **Link**: https://arxiv.org/abs/2103.01655 ; KAIST VIO dataset: https://github.com/zinuok/kaistviodataset
+- **Tier**: L1 (peer-reviewed conference, IROS-track preprint with public dataset)
+- **Publication Date**: arXiv 2021-03-02
+- **Timeliness Status**: ⚠️ Older than the 18-month Critical-novelty window, but **uniquely authoritative** for the specific question "do these VIO algorithms run on a Jetson?"; the included algorithms (VINS-Mono, VINS-Fusion, ROVIO, ALVIO, Stereo-MSCKF, Kimera, ORB-SLAM2-stereo) are all classical baselines whose runtime characteristics on ARM CPUs have not changed materially. Jetson hardware comparison (TX2 / Xavier NX / AGX Xavier) does NOT include Orin Nano — must extrapolate.
+- **Version Info**: Conference paper.
+- **Target Audience**: UAV state-estimation engineers picking a VIO for a Jetson companion
+- **Research Boundary Match**: **Strong match for the question**, partial for the hardware (no Orin Nano). KAIST VIO dataset is indoor mocap, not UAV-aerial-nadir — the *latency / CPU / memory* numbers transfer; the *accuracy* numbers do not transfer to our domain.
+- **Summary**: Comprehensive benchmark of 9 algorithms on TX2, Xavier NX, AGX Xavier: VINS-Mono, VINS-Fusion (CPU), VINS-Fusion-gpu, VINS-Fusion-imu, ROVIO, Stereo-MSCKF, ALVIO, Kimera, ORB-SLAM2-stereo. **Hard findings**: (a) on TX2, **VINS-Fusion (CPU) and VINS-Fusion-imu fail to run** due to insufficient memory and CPU performance — VINS-Fusion-gpu does run; (b) all algorithms except ROVIO show >100% CPU usage (multi-core utilisation, OK for our 6-core Orin Nano A78AE); (c) Kimera has the highest memory usage among VIO methods (numerous computations per keyframe), failure-prone on Xavier NX-class memory; (d) Stereo-MSCKF has the lowest memory among stereo VIOs; (e) ROVIO has the lowest CPU usage owing to its patch-tracking formulation. **Implication for project**: Jetson Orin Nano Super (8 GB shared, 6-core A78AE, Ampere GPU, 67 TOPS sparse INT8) is between Xavier NX and AGX Xavier in CPU performance and memory; algorithms passing on Xavier NX should pass on Orin Nano Super, but VINS-Fusion-imu's TX2 failure is a yellow-flag for memory pressure under co-resident C2/C3/C5 modules.
+- **Related Sub-question**: SQ3+SQ4 / C1 (VINS-Mono / VINS-Fusion / OpenVINS / Kimera / Stereo-MSCKF / ROVIO Jetson runtime evidence), SQ5 (resource-budget failure modes)
+
+
+### Source #47
+- **Title**: OKVIS2 — Realtime Scalable Visual-Inertial SLAM with Loop Closure (Leutenegger, ETH/Imperial/TUM Smart Robotics Lab)
+- **Link**: https://github.com/ethz-mrl/okvis2 ; arXiv: https://arxiv.org/abs/2202.09199 ; LICENSE: https://github.com/ethz-mrl/okvis2/blob/main/LICENSE
+- **Tier**: L1 (canonical implementation; arXiv 2022 by paper author)
+- **Publication Date**: original arXiv 2022; OKVIS2-X T-RO 2025 successor (Boche, Jung, Laina, Leutenegger — IEEE T-RO 2025, vol 41 pp 6064–6083, DOI 10.1109/TRO.2025.3619051; arXiv 2510.04612, Oct 2025). Repository last push 2026-03-17 (ethz-mrl/OKVIS2-X).
+- **Timeliness Status**: ✅ **Current.** Active development through 2026; OKVIS2-X is the most recent published VI-SLAM system in this class.
+- **Version Info**: ethz-mrl/okvis2 (core) and ethz-mrl/OKVIS2-X (multi-sensor extension with optional GNSS / LiDAR / dense depth).
+- **Target Audience**: Factor-graph VI-SLAM implementers; mid-large-scale loop-closure use cases
+- **Research Boundary Match**: **Full match** for monocular+IMU mode. OKVIS2 README + paper explicitly support mono and multi-camera VI configurations. OKVIS2-X adds GNSS fusion (relevant: VINS-Fusion-style GPS-when-available drop-in IS the project's eventual posture in non-spoofed regions).
+- **Summary**: Factor-graph VI-SLAM with bounded-size optimization. Innovation: pose-graph edges from marginalised observations can be "seamlessly turned back into observations" upon loop closure, reviving old landmarks and reprojection errors. Includes lightweight CNN segmentation for dynamic-region removal. OKVIS2-X (2025) generalises the core to fuse multi-camera + IMU + optional GNSS + LiDAR/depth — directly aligned with project's "VIO that may opportunistically fuse a non-spoofed GPS update" pattern and AC-NEW-2's spoof-promotion path. **License: 3-clause BSD (permissive)** — no copyleft / dual-use distribution friction. Note: GitHub UI shows "Other (NOASSERTION)" because of the standard BSD clause language pattern; the LICENSE file is canonical 3-clause BSD.
+- **Related Sub-question**: SQ3+SQ4 / C1 lead candidate (factor-graph + permissive license + active maintenance)
+
+
+### Source #48
+- **Title**: OKVIS2-X: Open Keyframe-based Visual-Inertial SLAM Configurable with Dense Depth or LiDAR, and GNSS (Boche, Jung, Laina, Leutenegger — TUM / ETH Zurich Smart Robotics Lab)
+- **Link**: https://github.com/ethz-mrl/OKVIS2-X ; arXiv: https://arxiv.org/abs/2510.04612 ; IEEE T-RO 2025 vol 41 pp 6064–6083 DOI 10.1109/TRO.2025.3619051
+- **Tier**: L1 (peer-reviewed IEEE Transactions on Robotics, Special Issue Visual SLAM 2025)
+- **Publication Date**: arXiv 2025-10-04; T-RO 2025 vol 41
+- **Timeliness Status**: ✅ Current (within 6-month Critical-novelty window)
+- **Version Info**: 295 stars; 38 forks; 2 contributors; created 2025-09-23, last push 2026-03-17. License: NOASSERTION on GitHub UI; per-paper license follows ethz-mrl convention (BSD-3 derived).
+- **Target Audience**: Multi-sensor SLAM researchers; large-scale VI-SLAM with optional GNSS/LiDAR
+- **Research Boundary Match**: **Strong match** — extends OKVIS2 monocular+IMU mode with optional GNSS fusion (Visual-Inertial SLAM with Tightly-Coupled Dropout-Tolerant GPS Fusion lineage from IROS 2022). Project's `MAV_CMD_SET_EKF_SOURCE_SET` switch + companion-side spoof-detection conceptually mirrors OKVIS2-X's "GPS as drop-out-tolerant signal".
+- **Summary**: Non-trivial extension of OKVIS2; submap-based volumetric occupancy mapping. Demonstrates that the OKVIS2 factor-graph backbone can absorb spoofing-aware GPS without re-architecting. Useful as architectural template for project's C5 estimator + C8 adapter integration. License: same as OKVIS2 (BSD-3-derived). Two named contributors (bochsim, SebsBarbas) actively pushing through Mar 2026.
+- **Related Sub-question**: SQ3+SQ4 / C1 (OKVIS2 lineage; VI-SLAM with optional GPS/LiDAR), SQ8 (GPS-fusion dropout-tolerant lineage)
+
+
+### Source #49
+- **Title**: Kimera-VIO — Visual Inertial Odometry with SLAM capabilities and 3D Mesh generation (MIT-SPARK)
+- **Link**: https://github.com/MIT-SPARK/Kimera-VIO ; LICENSE.BSD: https://github.com/MIT-SPARK/Kimera-VIO/blob/master/LICENSE.BSD
+- **Tier**: L1 (canonical implementation by MIT SPARK Lab)
+- **Publication Date**: original 2020 (Rosinol, Abate, Chang, Carlone — ICRA 2020); ongoing development through 2024–2025 issue threads (Dec 2024 / Feb 2025 ROS2 / mono-inertial discussion).
+- **Timeliness Status**: ✅ Active maintenance (recent issues / PRs through 2025).
+- **Version Info**: master-branch-only; LICENSE.BSD = BSD 2-Clause "Simplified".
+- **Target Audience**: VI-SLAM + mesh-mapping researchers
+- **Research Boundary Match**: **Partial.** Stereo+IMU is the primary supported configuration; mono+IMU is **optional but documented**. Kimera also produces 3D mesh and high-level semantic labels (relevant to neither C1 nor the project's bandwidth budget — overhead).
+- **Summary**: Frontend (image processing + IMU pre-integration) + Backend (factor-graph optimization in iSAM2 or GTSAM) + Mesher + Pose-Graph-Optimizer. **License: BSD 2-Clause (permissive)** — no dual-use distribution friction. **Penalty for project**: Source #46 KAIST benchmark found Kimera has highest memory usage among the VIOs tested (numerous computations per keyframe), and Kimera failed to fit on Xavier-NX-class memory under multi-process load. Mesh + semantic features are unused by the project — Kimera's overhead is unjustified vs OKVIS2 / OpenVINS for the project's narrow C1 mandate. **Status**: viable secondary fallback if OKVIS2 / VINS-Mono runtime issues arise; not a lead candidate due to overhead misfit.
+- **Related Sub-question**: SQ3+SQ4 / C1 secondary candidate (BSD-permissive but resource-heavy)
+
+
+### Source #50
+- **Title**: DROID-SLAM — Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras (princeton-vl, Teed & Deng)
+- **Link**: https://github.com/princeton-vl/droid-slam ; arXiv: https://arxiv.org/abs/2108.10869 ; NeurIPS 2021
+- **Tier**: L1 (canonical reference)
+- **Publication Date**: NeurIPS 2021; repository latest tagged baseline.
+- **Timeliness Status**: ✅ Foundational reference; DPV-SLAM (Source #51) is the lighter successor.
+- **Version Info**: master-branch-only.
+- **Target Audience**: Deep-learning-based VO/VSLAM researchers
+- **Research Boundary Match**: **Disqualified by hardware budget.** Inference requires ≥11 GB GPU VRAM per official README; project budget is 8 GB **shared CPU+GPU** on Jetson Orin Nano Super, leaving <8 GB for VO + VPR + matcher + estimator + cache co-resident. DROID-SLAM is also **monocular VO/SLAM, not VIO** — no native IMU fusion; metric scale recovery requires external scale alignment.
+- **Summary**: Recurrent dense bundle adjustment over a complete history of camera poses. State-of-the-art accuracy on TartanAir / EuRoC / TUM-RGBD at the cost of GPU memory. **Disqualified outright for C1 lead** by AC-4.2 (≤8 GB shared RAM) and the lack of IMU fusion that would require an additional ESKF/UKF wrapping. Kept as **reference baseline** to be cited as "what we cannot afford" in `solution_draft01`.
+- **Related Sub-question**: SQ3+SQ4 / C1 disqualified candidate
+
+
+### Source #51
+- **Title**: DPVO — Deep Patch Visual Odometry (princeton-vl, Teed, Lipson, Deng) + DPV-SLAM (Lipson, Teed, Deng — ECCV 2024)
+- **Link**: https://github.com/princeton-vl/DPVO ; LICENSE: https://github.com/princeton-vl/DPVO/blob/main/LICENSE ; ECCV 2024 paper: https://www.ecva.net/papers/eccv_2024/papers_ECCV/papers/00272.pdf
+- **Tier**: L1 (canonical implementation; NeurIPS 2023 + ECCV 2024)
+- **Publication Date**: NeurIPS 2023 (DPVO); ECCV 2024 (DPV-SLAM); repository last update 2024-10-12.
+- **Timeliness Status**: ⚠️ Borderline. ~19 months since last code update; ECCV-2024 publication of DPV-SLAM keeps the algorithm class within the 6-month claim window for the SLAM successor.
+- **Version Info**: 989 stars; primary languages C++ / Python / CUDA. **License: MIT (permissive)** — no dual-use distribution friction.
+- **Target Audience**: Deep-learning VO/SLAM with reduced memory footprint
+- **Research Boundary Match**: **Partial.** DPVO is **monocular VO only — no IMU fusion**. Output pose is in arbitrary scale (no metric scale recovery). To be a viable C1 candidate the project must wrap DPVO with an external IMU+scale-fusion stage (loosely-coupled ESKF / VIO-fusion module). This makes DPVO **not a drop-in C1** like VINS-Mono / OpenVINS / OKVIS2; it is a **VO module that needs a separate VIO wrapper**.
+- **Summary**: Sparse patch tracking + differentiable bundle adjustment back end. Outperforms DROID-SLAM on TartanAir / EuRoC ATE while using ~1/3 of DROID-SLAM's GPU memory (DROID-SLAM: 8.7 GB VO mode vs DPVO: ~3 GB). DPV-SLAM (Lipson, Teed, Deng — ECCV 2024) adds full SLAM capability with 4–5 GB GPU usage. **Jetson runtime evidence**: indirect via DPVO-QAT++ (Source #52) — peak reserved memory 1.02 GB on RTX 4060 (8 GB) after INT8 fake-quant + custom CUDA kernel fusion; not directly tested on Jetson Orin Nano. **Status for C1**: pure-VO candidate (must be paired with separate IMU integration to deliver metric scale + attitude); would not satisfy "monocular VIO" gate alone, but viable as the *VO half* of a hybrid C1+C5 design.
+- **Related Sub-question**: SQ3+SQ4 / C1 conditional candidate (VO not VIO; needs external IMU wrapper)
+
+
+### Source #52
+- **Title**: DPVO-QAT++: Heterogeneous QAT and CUDA Kernel Fusion for High-Performance Deep Patch Visual Odometry (Cheng Liao)
+- **Link**: https://arxiv.org/abs/2511.12653 ; project HTML: https://arxiv.org/html/2511.12653
+- **Tier**: L2 (single-author preprint, code partially released; no peer-review yet)
+- **Publication Date**: arXiv 2025-11-16 (within 6-month Critical-novelty window)
+- **Timeliness Status**: ✅ Current
+- **Version Info**: arXiv preprint; code & weights released for QAT-only and fused-CUDA variants.
+- **Target Audience**: Embedded-platform DPVO deployers
+- **Research Boundary Match**: **Partial.** Hardware tested = RTX 4060 (8 GB) + Intel Core Ultra 5-125H + 32 GB RAM — desktop GPU, NOT Jetson Orin Nano. Direct extrapolation requires Jetson MVE; Orin Nano Super's Ampere GPU is architecturally similar but smaller than RTX 4060.
+- **Summary**: Quantization-Aware Training framework for DPVO with fused CUDA kernels. Reduces peak GPU memory from 1.94 GB → 1.02 GB (-47%) on a representative TartanAir sequence; +34.6% median FPS on TartanAir, +26.7% on EuRoC; -22.8 ms / -19.7 ms median P99 tail latency on TartanAir / EuRoC respectively. Heterogeneous precision: front-end pseudo-quantization (FP16/FP32 with INT8 simulation) + FP32 back-end geometric solver. **Implication for project**: shows DPVO has a documented Jetson-suitable footprint **path** but not a Jetson-Orin-Nano measurement. ATE accuracy comparable to baseline DPVO across 32 TartanAir + 11 EuRoC validation sequences. Notable: requires a teacher-student distillation training pipeline before deployment — adds operational complexity vs classical VINS-* / OpenVINS / OKVIS2.
+- **Related Sub-question**: SQ3+SQ4 / C1 supporting evidence for DPVO embedded feasibility
+
+
+### Source #53
+- **Title**: Pure VO baseline — KLT optical flow + 5-point essential matrix or homography RANSAC (OpenCV reference)
+- **Link**: https://docs.opencv.org/4.x/d4/dee/tutorial_optical_flow.html ; representative public implementation: https://github.com/alishobeiri/Monocular-Video-Odometery (MIT, 2018) ; tutorial reference: https://zxh.me/posts/2022-12-19-monocular-visual-odometry/
+- **Tier**: L1 (OpenCV official documentation) + L2 (representative public implementations)
+- **Publication Date**: OpenCV docs continuously updated; tutorial 2022-12; reference implementation 2018 (algorithmic class is foundational, no time window per Step 0.5)
+- **Timeliness Status**: ✅ Foundational baseline (no time window).
+- **Version Info**: OpenCV `cv::calcOpticalFlowPyrLK` (KLT) + `cv::findEssentialMat` (5-point Nister) or `cv::findHomography` with RANSAC.
+- **Target Audience**: Implementers needing a transparent low-complexity fallback
+- **Research Boundary Match**: **Full match for the simple-baseline candidate.** Suits planar nadir-down UAV at altitude (Ukrainian steppe is ~planar at 1 km AGL — homography is geometrically appropriate; for non-planar relief the essential matrix path is more appropriate but adds scale-recovery work).
+- **Summary**: Established classical pipeline: Shi-Tomasi or FAST corner detection → KLT pyramidal optical flow tracking → 5-point essential matrix or homography RANSAC → relative pose with arbitrary scale (must be metric-scale-aligned via IMU integration externally). Reference implementations widely available in OpenCV samples and pedagogical repos. **Status**: candidate as the project's `Simple baseline / known-runnable / known-failure-mode` C1 option per Component Option Breadth rule. Not a lead, but mandatory fallback presence per the research engine's "include at least one simple baseline" rule.
+- **Related Sub-question**: SQ3+SQ4 / C1 simple-baseline candidate
+
+
+### Source #54
+- **Title**: OpenVINS — `context7` per-mode capability lookup (`/rpng/open_vins`, master)
+- **Link**: context7 query against `/rpng/open_vins`, accessed 2026-05-08; canonical doc references returned: `https://github.com/rpng/open_vins/blob/master/docs/gs-tutorial.dox`, `https://github.com/rpng/open_vins/blob/master/docs/gs-datasets.dox`, `https://github.com/rpng/open_vins/blob/master/docs/gs-calibration.dox`, `https://github.com/rpng/open_vins/blob/master/docs/propagation-analytical.dox`
+- **Tier**: L1 (project-official documentation reachable via the project's documentation generator)
+- **Publication Date**: live docs (master, accessed 2026-05-08)
+- **Timeliness Status**: ✅ Within Critical-novelty window (active master + community evidence through 2025–2026)
+- **Version Info**: master HEAD at access time (no tagged release for ROS 2 path; ROS 1 / ROS 2 build paths both documented)
+- **Target Audience**: System architects + C1 implementer
+- **Research Boundary Match**: **Full match** for monocular + IMU mode. The `subscribe.launch.py` ROS 2 launch script (and its ROS 1 sibling) declare `use_stereo` and `max_cameras` as DeclareLaunchArguments — setting `use_stereo:=false max_cameras:=1` selects monocular operation; `config:=` selects an estimator-config directory (`euroc_mav`, `tum_vi`, `rpng_aruco`, …). KALIBR + RPNG IMU intrinsic calibration models are both documented in `propagation-analytical.dox` with the corresponding state-vector composition.
+- **Summary**: Confirms documentary evidence for OpenVINS' three sensor configurations exposed at the launch layer (mono / stereo / multi-camera), all with IMU mandatory; confirms the project's pinned mode (`use_stereo:=false max_cameras:=1`) is a first-class launch configuration that requires no source patch. Confirms that estimator config files in `ov_msckf/config/<dataset>/estimator_config.yaml` are the parameter-tuning surface and that supported IMU intrinsic models include both KALIBR and RPNG. **Open**: `context7` Disqualifier-Probe query did not surface explicit per-mode latency/memory limits or sub-20-Hz validation evidence; those constraints carry into the Jetson-Orin-Nano-Super hardware MVE (D-C1-2 deferred phase).
+- **Related Sub-question**: SQ3+SQ4 / C1 — OpenVINS per-mode API capability verification (Mandatory `context7` lookup per Per-Mode API Capability Verification rule)
+
+
+### Source #55
+- **Title**: VINS-Mono — official README + VINS-Fusion `context7` per-mode capability lookup (`/hkust-aerial-robotics/vins-fusion`, master) [cross-source documentary evidence for the mono+IMU mode shared with VINS-Mono]
+- **Link**: VINS-Mono README — https://raw.githubusercontent.com/HKUST-Aerial-Robotics/VINS-Mono/master/README.md (accessed 2026-05-08); VINS-Fusion docs — context7 query against `/hkust-aerial-robotics/vins-fusion`, accessed 2026-05-08, canonical reference returned: https://github.com/hkust-aerial-robotics/vins-fusion/blob/master/README.md
+- **Tier**: L1 (project-official READMEs of both repos)
+- **Publication Date**: VINS-Mono README — 2019-01-11 last major revision (master-branch only, no tagged releases); VINS-Fusion docs — live (master, accessed 2026-05-08)
+- **Timeliness Status**: ⚠️ borderline (per Step 0.5 timeliness — VINS-Mono master last meaningful commit 2024-02-25 / 2024-05-23; older than the 18-month preferred window for live API behaviour, but the algorithm class remains the canonical mono+IMU sliding-window VIO referenced by 2025 community work — see Fact #36)
+- **Version Info**: VINS-Mono master HEAD; depends on Ceres v1.14.0 (versions ≥2.0.0 have build issues per README). VINS-Fusion master HEAD has `euroc_mono_imu_config.yaml` as a first-class config.
+- **Target Audience**: System architects + C1 implementer
+- **Research Boundary Match**: **Full match** for the project's pinned mode (mono + IMU). VINS-Mono is single-mode by construction — "real-time SLAM framework for **Monocular Visual-Inertial Systems**" — the project's pinned mode is the only mode the project will use the binary in. VINS-Fusion `euroc_mono_imu_config.yaml` is the documentary cross-source evidence that the algorithmic mono+IMU path remains a first-class configuration in the same authors' active fork.
+- **Summary**: Confirms VINS-Mono = monocular + IMU only (single mode); ROS Kinetic / Ubuntu 16.04 reference build; pinhole + MEI camera models supported; rolling-shutter support with calibrated reprojection error <0.5 px; online camera-IMU extrinsic + temporal calibration; loop closure via DBoW2; pose-graph reuse and map merge supported. **Critical recommended-input bound**: README §5.1 — *"The image should exceed 20Hz and IMU should exceed 100Hz."* — the project's nav cam target is 3 fps; this is a documentary signal that VIO performance below the recommended frame rate is not validated by the upstream authors. License: GPLv3 (confirmed in README §8). **Cross-source note**: VINS-Fusion `euroc_mono_imu_config.yaml` is named explicitly in `context7` results and uses the same algorithmic core; treat as evidence for VINS-Mono's mono+IMU mode while honouring the per-mode rule that VINS-Fusion's mono+IMU mode is a separately-cataloged candidate (Fact #29).
+- **Related Sub-question**: SQ3+SQ4 / C1 — VINS-Mono per-mode API capability verification (Mandatory `context7` lookup per Per-Mode API Capability Verification rule, with cross-source documentary evidence from VINS-Fusion since VINS-Mono itself is not indexed in `context7`)
+
+
+### Source #56
+- **Title**: OKVIS2 — official README (`smartroboticslab/okvis2`, main)
+- **Link**: https://raw.githubusercontent.com/smartroboticslab/okvis2/main/README.md (accessed 2026-05-08); papers cited in README: arXiv:2202.09199 (Leutenegger 2022), IJRR 2015, RSS 2013
+- **Tier**: L1 (project-official README; arXiv canonical paper)
+- **Publication Date**: README live; canonical paper 2022-02; OKVIS2 master last push within the Critical-novelty window (per Fact #36 timeliness audit, OKVIS2-X 2026-03-17 push confirms active)
+- **Timeliness Status**: ✅ Fully within Critical-novelty window
+- **Version Info**: OKVIS2 main HEAD; cmake build with optional ROS 2 wrapping (`BUILD_ROS2=ON`); optional sky-segmentation CNN via LibTorch (`USE_NN=OFF` to disable)
+- **Target Audience**: System architects + C1 implementer + Step-7.5 reviewer
+- **Research Boundary Match**: **Full match** for the project's pinned mode (mono + IMU). README confirms multi-camera support (camera frames `C_i` for the i-th camera) plus IMU mandatory; mono operation is a documented configuration via the example apps (`okvis_app_synchronous`, `okvis_app_realsense`). OKVIS2-X is the GNSS-fusion extension (T-RO 2025) that aligns architecturally with the project's spoof-promotion path.
+- **Summary**: Confirms OKVIS2 = keyframe-based VI-SLAM (factor-graph backbone with loop closure); BSD-3 license (no copyleft); coordinate-frame contract (`W` world, `C_i` cameras, `S` IMU, `B` body); state representation (`T_WS` pose + velocity + gyro/accel biases); two-callback API (`setOptimisedGraphCallback` for batch updates incl. loop closure + `setImuCallback` for high-rate prediction). Calibration prerequisites: camera intrinsics + camera-IMU extrinsics + IMU noise parameters + tight time sync (Kalibr toolchain explicitly recommended). Optional LibTorch sky-segmentation CNN can be disabled (`USE_NN=OFF`) to remove a major Jetson dependency. ROS 2 build path (`BUILD_ROS2=ON`) with `okvis_node_realsense.launch.xml`, `okvis_node_realsense_publisher.launch.xml`, `okvis_node_subscriber.launch.xml`, `okvis_node_synchronous.launch.xml`. **Health warning** in README: poor calibration → poor results; this is shared with all VI candidates but is more strongly emphasised in OKVIS2 docs. **Open**: README does not state explicit minimum frame rate (cf. VINS-Mono's documented 20 Hz minimum) — keyframe-based selection generally tolerates lower input frame rates than sliding-window optimisation; this needs explicit Jetson MVE validation at 3 fps.
+- **Related Sub-question**: SQ3+SQ4 / C1 — OKVIS2 per-mode API capability verification (Mandatory `context7` lookup per Per-Mode API Capability Verification rule, with WebFetch fallback to official README since `context7` returned no match)

_docs/00_research/01_source_registry/C4_pose_estimation.md

@@ -0,0 +1,88 @@
+# Source Registry — C4 — Pose estimation (PnP + RANSAC + LM) candidates
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation).
+> Critical-novelty sensitivity per Step 0.5 in `../00_question_decomposition.md`. Time windows applied:
+> - **Lead-candidate / SOTA claims**: prefer sources within last 6 months; up to 18 months if older is the official authority.
+> - **Library/SDK API behaviour**: must reflect the currently shipped version at search time (`context7` mandatory per lead candidate).
+> - **Established baselines** (KLT, RANSAC, EKF, ORB, SIFT, GTSAM): no time window.
+>
+> This file replaces a section of the previous monolithic `01_source_registry.md`. See `00_summary.md` for the full category index. Investigation order is tracked in `../00_question_decomposition.md` and the cross-category Investigation Status table in `00_summary.md`.
+
+---
+
+### Source #82
+- **Title**: OpenCV canonical implementation — `opencv/opencv` (Open Source Computer Vision Library) GitHub repository metadata via GitHub API + LICENSE — **Apache-2.0** (`license.spdx_id: "Apache-2.0"`); 87385 stars + 56554 forks + 2606 subscribers + 2732 open issues; created 2012-07-19; **last pushed 2026-05-08T07:00:03Z = TODAY at access time** (daily-active maintenance); default branch `4.x`; size 555 GB; topics include `c-plus-plus, computer-vision, deep-learning, image-processing, opencv`
+- **Link**: GitHub API metadata https://api.github.com/repos/opencv/opencv (accessed 2026-05-08; `license.spdx_id: "Apache-2.0"` confirmed); canonical repo https://github.com/opencv/opencv ; canonical website https://opencv.org ; LICENSE file https://raw.githubusercontent.com/opencv/opencv/4.x/LICENSE (Apache License 2.0 standard text)
+- **Tier**: L1 (project-official codebase by the OpenCV organization; canonical reference computer-vision library used by every modern computer-vision deployment as the de-facto industry-standard classical-CV foundation; cited by every C-row component's deployment guide; canonical solvePnPRansac is the industry-standard reference RANSAC-PnP implementation that every modern alternative [OpenGV, GTSAM-PnP, Theia, Ceres-only] compares against in its own documentation)
+- **Publication Date**: original 2000 (Intel) → open-source release 2006 (Willow Garage) → OpenCV.org foundation 2020 → canonical 4.x branch active continuous development; access date 2026-05-08; daily commits to `4.x` branch
+- **Timeliness Status**: ✅ Within Established-baseline-reference window (2000+ — established competitive ground for classical computer-vision + RANSAC-PnP reference; Established-competitive-mandatory-baseline exemption applies — `cv::solvePnPRansac` is the **canonical RANSAC-PnP reference baseline** that defines the mandatory-simple-baseline role for the C4 row per the engine Component Option Breadth rule, structurally analogous to NetVLAD's role in C2 row + SuperGlue+SuperPoint's role in C3 row)
+- **Version Info**: 4.14.0-pre at access time (default branch `4.x` = next-major-release rolling-development branch; current stable release 4.10.0 from late 2025 at access date — 4.x is the project's pinned major version per Source #83 documentation footer "Generated on Fri May 8 2026 04:21:44 for OpenCV by 1.12.0"); JetPack 6 ships canonical `libopencv_calib3d.so` for ARM Cortex-A78AE = the project's pinned Jetson Orin Nano Super deployment runtime
+- **Target Audience**: System architects + C4 implementer + Step-7.5 reviewer + license-posture decision-maker (D-C1-1 — clean Apache-2.0) + C7 (Jetson runtime) implementer (canonical OpenCV is shipped with JetPack 6 distribution)
+- **Research Boundary Match**: **Full match** for the project's pinned C4 mandatory-simple-baseline mode (per-frame pose-from-correspondences via classical RANSAC-PnP with paired Levenberg-Marquardt refinement). The canonical `opencv/opencv` library ships everything needed for C4 deployment: `cv::solvePnPRansac` two function signatures (classical + USAC variant), nine `SolvePnPMethod` enum values, paired `cv::solvePnPRefineLM` LM refinement + alternate `cv::solvePnPRefineVVS` Gauss-Newton SO(3) refinement, paired `cv::solvePnPGeneric` for multi-solution + per-solution reprojection-error reporting, `cv::projectPoints` Jacobian for D-C4-2 post-hoc covariance recovery. **N/A for the project's domain caveat** — OpenCV solvePnPRansac is a classical algorithm with no training data; D-C2-1 retrain decision is irrelevant for OpenCV solvePnPRansac
+- **Summary**: OpenCV is the canonical industry-standard open-source computer vision library; the calib3d module ships `cv::solvePnPRansac` as the canonical RANSAC-PnP reference implementation. **CRITICAL LICENSE FINDING**: Apache-2.0 (`license.spdx_id: "Apache-2.0"`) — permissive, BSD/permissive license track on the C4 mandatory-simple-baseline; **deployment-ready under every D-C1-1 license-posture choice** with the cleanest license-compliance story tied with cvg/LightGlue + DISK + XFeat. **Daily-active maintenance**: last pushed 2026-05-08 (TODAY at access time) — among the most actively-maintained C-row references across all components evaluated. **Industry-standard reference status**: 87385 stars + 56554 forks + 2606 subscribers — the dominant industry-standard reference implementation that every modern C4 alternative (OpenGV, GTSAM-PnP, Theia, Ceres-only) compares against in its own documentation. **JetPack 6 canonical distribution**: canonical OpenCV is shipped with JetPack 6 distribution, providing zero-effort deployment for the project's pinned Jetson Orin Nano Super runtime
+- **Related Sub-question**: SQ3+SQ4 / C4 — OpenCV solvePnPRansac per-mode API capability verification (Mandatory `context7` lookup MCP-validation-error + WebFetch fallback PASS per Per-Mode rule item 2; cross-validated against canonical GitHub API license metadata WebFetch + canonical OpenCV calib3d module documentation [Source #83]); **D-C1-1 license-posture compliance**: clean Apache-2.0 throughout; **Mandatory-simple-baseline role per engine Component Option Breadth rule** confirmed; **JetPack 6 canonical distribution** documented
+
+
+### Source #83
+- **Title**: OpenCV 4.x calib3d module canonical documentation — group `cv::calib3d` (Camera Calibration and 3D Reconstruction) at `https://docs.opencv.org/4.x/d9/d0c/group__calib3d.html` + Perspective-n-Point (PnP) pose computation tutorial at `https://docs.opencv.org/4.x/d5/d1f/calib3d_solvePnP.html`; `cv::solvePnPRansac` two function signatures (classical with `iterationsCount=100, reprojectionError=8.0, confidence=0.99, flags=SOLVEPNP_ITERATIVE` defaults + USAC variant with `UsacParams` and `cameraMatrix` as `InputOutputArray` for focal-length refinement); Python bindings; `cv::SolvePnPMethod` enum 9 values; `cv::solvePnPRefineLM` + alternate `cv::solvePnPRefineVVS`; `cv::solvePnPGeneric` for multi-solution + per-solution reprojection-error reporting; USAC RANSAC-method enum 7 modern variants
+- **Link**: calib3d module documentation https://docs.opencv.org/4.x/d9/d0c/group__calib3d.html (accessed 2026-05-08); PnP tutorial page https://docs.opencv.org/4.x/d5/d1f/calib3d_solvePnP.html (accessed 2026-05-08); both pages footer-stamped "Generated on Fri May 8 2026 04:21:44 for OpenCV by 1.12.0" — fresh canonical documentation at the project's evaluation time
+- **Tier**: L1 (canonical project-official documentation by the OpenCV organization; the canonical reference for the `cv::solvePnPRansac` function signature, parameter defaults, paired refinement variants, minimal-solver enum values, and structural caveats; auto-generated by Doxygen 1.12.0 from canonical opencv/opencv source code at `4.x` branch)
+- **Publication Date**: rolling Doxygen documentation auto-regenerated on every push to `4.x` branch; access date 2026-05-08 04:21:44 page-generation timestamp
+- **Timeliness Status**: ✅ Within Established-baseline-reference window (rolling Doxygen documentation; the canonical reference for `cv::solvePnPRansac` API surface at the project's evaluation time)
+- **Version Info**: 4.14.0-pre at access time (default branch `4.x` = next-major-release rolling-development branch). **Mode-enumeration query (1/3) — context7 MCP-validation-error + WebFetch fallback PASS** — `context7 resolve-library-id` returned MCP validation errors (parameter schema mismatch on both `query` and `libraryName` argument names — context7 server expects different argument shape than provided); per Per-Mode API Capability Verification rule item 2, fall-back to official-docs WebFetch on the canonical OpenCV calib3d module documentation + PnP tutorial page was used (this Source #83). **Nine `SolvePnPMethod` enum values documented** at line 243 of the calib3d.html: `SOLVEPNP_ITERATIVE=0` (default; iterative LM-based on top of EPNP minimal-solver result), `SOLVEPNP_EPNP=1` (Efficient Perspective-n-Point [Lepetit et al. IJCV 2009]; canonical default for ≥4 non-planar correspondences), `SOLVEPNP_P3P=2` (Revisiting the P3P Problem [Ding et al. 2023]; minimal-solver for exactly-3 correspondences with up to 4 solutions), `SOLVEPNP_DLS=3` (**BROKEN per explicit docstring "Broken implementation. Using this flag will fallback to EPnP"** — Direct Least-Squares method [Hesch & Roumeliotis 2011] originally), `SOLVEPNP_UPNP=4` (**BROKEN per explicit docstring "Broken implementation. Using this flag will fallback to EPnP"** — Exhaustive Linearization for Robust Camera Pose and Focal Length Estimation [Penate-Sanchez et al. 2013] originally), `SOLVEPNP_AP3P=5` (Algebraic P3P [Ke & Roumeliotis CVPR 2017]), `SOLVEPNP_IPPE=6` (Infinitesimal Plane-Based Pose Estimation [Collins & Bartoli ECCV 2014]; **planar-only — object points must be coplanar — directly relevant to project's D-C4-1 = 4-DoF flat-earth lift recommendation**), `SOLVEPNP_IPPE_SQUARE=7` (special-case IPPE for marker pose with 4 fixed-pattern points), `SOLVEPNP_SQPNP=8` (SQPnP: A Consistently Fast and Globally Optimal Solution [Terzakis & Lourakis ECCV 2020]; **modern globally-optimal alternate without planarity restriction — second-recommended fallback if D-C4-1 chooses 6-DoF DSM lift**). **`cv::solvePnPRansac` classical signature** at line 3211 of calib3d.html: `bool solvePnPRansac(InputArray objectPoints, InputArray imagePoints, InputArray cameraMatrix, InputArray distCoeffs, OutputArray rvec, OutputArray tvec, bool useExtrinsicGuess=false, int iterationsCount=100, float reprojectionError=8.0, double confidence=0.99, OutputArray inliers=noArray(), int flags=SOLVEPNP_ITERATIVE)` — Python `cv.solvePnPRansac(objectPoints, imagePoints, cameraMatrix, distCoeffs[, rvec[, tvec[, useExtrinsicGuess[, iterationsCount[, reprojectionError[, confidence[, inliers[, flags]]]]]]]]) -> retval, rvec, tvec, inliers`. **`cv::solvePnPRansac` USAC variant signature** at line 3261: `bool solvePnPRansac(InputArray objectPoints, InputArray imagePoints, InputOutputArray cameraMatrix, InputArray distCoeffs, OutputArray rvec, OutputArray tvec, OutputArray inliers, const UsacParams& params=UsacParams())` — Python `cv.solvePnPRansac(objectPoints, imagePoints, cameraMatrix, distCoeffs[, rvec[, tvec[, inliers[, params]]]]) -> retval, cameraMatrix, rvec, tvec, inliers`; note `cameraMatrix` is `InputOutputArray` in the USAC variant, allowing focal-length refinement during the RANSAC loop. **`cv::solvePnPRefineLM`** at line 3268: canonical default `TermCriteria(EPS+COUNT, 20, FLT_EPSILON)`. **CRITICAL CAVEAT** documented at the PnP-tutorial page: "the current implementation computes the rotation update as a perturbation and not on SO(3)" — minor structural caveat; alternate `cv::solvePnPRefineVVS` at line 3289 uses Gauss-Newton with rotation update via exponential map on SO(3) (preferred for high-accuracy aerial pose-from-correspondences). **`cv::solvePnPGeneric`** at line 370: returns multiple candidate solutions sorted by reprojection error + an `OutputArray reprojectionError` per-solution. **Default minimal-sample-set method** at line 3256: "The default method used to estimate the camera pose for the Minimal Sample Sets step is `SOLVEPNP_EPNP`. Exceptions are: if you choose `SOLVEPNP_P3P` or `SOLVEPNP_AP3P`, these methods will be used; if the number of input points is equal to 4, `SOLVEPNP_P3P` is used." **USAC RANSAC-method enumeration** at the calib3d.html anonymous-enum block: canonical RANSAC, LMEDS, RHO, **USAC_DEFAULT, USAC_PARALLEL, USAC_FM_8PTS, USAC_FAST, USAC_ACCURATE, USAC_PROSAC, USAC_MAGSAC** — modern USAC variants (Barath et al. CVPR 2019 + ICCV 2019 MAGSAC++) provide higher inlier-recovery rate than vanilla RANSAC at the same iteration budget; **USAC_MAGSAC is the canonical sigma-consensus modern alternative to vanilla RANSAC** with no fixed inlier threshold
+- **Target Audience**: System architects + C4 implementer + Step-7.5 reviewer + Plan-phase architect (mandatory-simple-baseline role documentation for engine Component Option Breadth rule compliance + D-C4-1 2D-3D-lift architectural decision carry-forward + D-C4-2 NEW covariance-recovery-strategy gate)
+- **Research Boundary Match**: **Full match** for the C4 row's pinned mode (per-frame pose-from-correspondences contract on Jetson Orin Nano Super; inputs = up to 1024 3D-2D correspondences from C3's 2D-2D + D-C4-1's 2D→3D lift + camera intrinsic + distortion; outputs = 6-DoF camera pose + per-correspondence inlier mask + reprojection error + RANSAC iter count + 6×6 covariance via D-C4-2). The canonical OpenCV calib3d module documentation provides the complete API surface for the project's pinned mode: two function signatures, nine minimal-solver enum values, paired LM + Gauss-Newton SO(3) refinement, paired multi-solution reporting with reprojection error, USAC RANSAC-method enumeration with 7 modern variants. **CRITICAL contract finding**: the documented signature requires `objectPoints` Nx3 1-channel + `imagePoints` Nx2 1-channel — **3D-2D, not 2D-2D**; the project must perform a 2D→3D lift on C3's satellite-tile-side 2D pixels via D-C4-1's 4-DoF flat-earth lift recommendation (project default) before calling solvePnPRansac. **CRITICAL covariance finding**: the documented signature returns `retval, rvec, tvec, inliers` only — **NO direct 6×6 covariance output**; AC-NEW-4 covariance-honesty contract requires D-C4-2 NEW Plan-phase decision for covariance-recovery-strategy
+- **Summary**: The canonical OpenCV 4.x calib3d module documentation is the definitive reference for `cv::solvePnPRansac` API surface, parameter defaults, paired refinement variants, minimal-solver enum values, and structural caveats. Two function signatures (classical + USAC variant), nine `SolvePnPMethod` enum values (4 valid for general project use + 2 special-case + 1 ITERATIVE default + 2 BROKEN-fallback-to-EPNP), paired `cv::solvePnPRefineLM` (LM with rotation update as perturbation, NOT on SO(3)) + alternate `cv::solvePnPRefineVVS` (Gauss-Newton on SO(3) via exponential map) refinement, paired `cv::solvePnPGeneric` for multi-solution + per-solution reprojection-error reporting, USAC RANSAC-method enumeration with 7 modern variants (USAC_DEFAULT, USAC_PARALLEL, USAC_FM_8PTS, USAC_FAST, USAC_ACCURATE, USAC_PROSAC, USAC_MAGSAC). **CRITICAL findings for the C4 row**: (i) **3D-2D INPUT CONTRACT, NOT 2D-2D** — solvePnPRansac requires Nx3 objectPoints + Nx2 imagePoints; project must perform 2D→3D lift via D-C4-1's locked-in 4-DoF flat-earth lift recommendation before invocation; (ii) **NO DIRECT 6×6 COVARIANCE OUTPUT** — AC-NEW-4 covariance-honesty contract requires D-C4-2 NEW Plan-phase decision for covariance-recovery-strategy; (iii) **TWO MINIMAL-SOLVER ENUM VALUES BROKEN** — SOLVEPNP_DLS + SOLVEPNP_UPNP fall back to EPNP per explicit docstring; valid set is `EPNP / AP3P / IPPE / SQPNP` plus 2 special-case (`P3P` for exactly-3; `IPPE_SQUARE` for 4-fixed-pattern markers) plus `ITERATIVE` default; (iv) **`cv::solvePnPRefineLM` ROTATION UPDATE NOT ON SO(3)** — minor caveat; alternate `cv::solvePnPRefineVVS` is the SO(3)-correct refiner. Canonical default minimal-sample-set method is `SOLVEPNP_EPNP`; recommended pairing for D-C4-1 = 4-DoF flat-earth lift is `SOLVEPNP_IPPE` (planar-scene minimal-solver designed for coplanar object points) with `SOLVEPNP_SQPNP` as the modern globally-optimal fallback
+- **Related Sub-question**: SQ3+SQ4 / C4 — OpenCV solvePnPRansac per-mode API capability verification (cross-source verification of canonical API documentation + structural caveats + minimal-solver enum + paired refinement variants); **D-C4-2 NEW Plan-phase decision raised** for covariance-recovery-strategy; **D-C4-1 carry-forward REINFORCED** by the 3D-2D-input-contract finding (applies to all C4 candidates, not unique to OpenCV); cross-cite to Fact #20 + #21 closures from C2 row (canonical PnP+RANSAC+LM reference pipeline shape feeds AC-NEW-4 covariance-honesty contract)
+
+
+### Source #84
+- **Title**: OpenGV canonical implementation — `laurentkneip/opengv` (A library for solving calibrated central and non-central geometric vision problems) GitHub repository metadata via GitHub API + License.txt — **BSD-3-Clause-equivalent boilerplate** ("Author: Laurent Kneip, ANU. All rights reserved." with three numbered redistribution conditions including non-endorsement clause; **GitHub API license SPDX detector reports `license.spdx_id: "NOASSERTION"`** because the License.txt file does NOT use the canonical Open Source Initiative BSD-3-Clause boilerplate text — verified by direct WebFetch of `https://raw.githubusercontent.com/laurentkneip/opengv/master/License.txt`); 1109 stars + 358 forks + 66 subscribers + 58 open issues; created 2013-08-10; **last pushed 2023-06-07T18:14:14Z = ~2 years 11 months stale at access time 2026-05-08** (CRITICAL maintenance finding); default branch `master`; size 7790 KB; description "OpenGV is a collection of computer vision methods for solving geometric vision problems. It is hosted and maintained by the Mobile Perception Lab of ShanghaiTech."
+- **Link**: GitHub API metadata https://api.github.com/repos/laurentkneip/opengv (accessed 2026-05-08); canonical repo https://github.com/laurentkneip/opengv ; License.txt https://raw.githubusercontent.com/laurentkneip/opengv/master/License.txt (BSD-3-Clause-equivalent boilerplate verified via WebFetch); canonical Doxygen documentation portal https://laurentkneip.github.io/opengv/
+- **Tier**: L1 (project-official codebase by Laurent Kneip + ShanghaiTech Mobile Perception Lab; canonical reference for non-OpenCV PnP solvers including p3p_kneip [Kneip et al. CVPR 2011], p3p_gao [Gao et al. PAMI 2003], UPnP [Kneip et al. ECCV 2014], gpnp [Kneip 2014 generalized PnP], gp3p [generalized 3-point]; cited by every modern multi-camera + central-camera + relative-pose paper since 2014; field-standard for non-trivial PnP variants beyond OpenCV's `cv::solvePnPRansac` coverage)
+- **Publication Date**: original 2013-08-10 → continuous development 2013-2018 → maintenance gap 2018-2023 → last pushed 2023-06-07; access date 2026-05-08; **Doxygen documentation portal generation timestamp "Generated on Mon Jan 8 2018 21:43:04 for OpenGV by 1.8.11" — documentation page is 8.3 years old at access time**
+- **Timeliness Status**: ⚠️ Within Established-baseline-reference window (2013+ — established competitive ground for non-OpenCV PnP minimal solvers + generalized-camera support) but **with CRITICAL ~3-year maintenance staleness caveat** — Established-competitive-mandatory-baseline exemption applies (OpenGV is the canonical reference for non-trivial PnP variants beyond OpenCV) but Plan-phase decision-maker MUST account for: (i) no security patches since 2023; (ii) no Eigen 3.4+ compatibility patches; (iii) no JetPack 6 + ARM Cortex-A78AE compilation testing in upstream CI; (iv) ShanghaiTech Mobile Perception Lab's claim of active maintenance is contradicted by the GitHub commit history at access time
+- **Version Info**: master branch at git commit ea7c66f5e (last commit 2023-06-07T18:14:14Z); no version tags, no releases. **Mode-enumeration query (1/3) — context7 NOT INDEXED + WebFetch fallback PASS** — `context7 resolve-library-id` returned only OpenCV variants for the OpenGV query (top-5 results were `/websites/opencv_4_x` + `/websites/opencv_4_6_0` + `/opencv/opencv` + `/opencv/opencv-python` + `/websites/opencv_5_0_0-alpha` — all unrelated to OpenGV); per Per-Mode API Capability Verification rule item 2, fall-back to official-docs WebFetch on canonical Doxygen portal `laurentkneip.github.io/opengv/page_how_to_use.html` was used (this Source #85 below + License.txt verification on this Source #84). **Absolute pose minimal solvers documented** via Source #85 §"Central absolute pose": `absolute_pose::p2p` (with known rotation), `absolute_pose::p3p_kneip` [Kneip CVPR 2011], `absolute_pose::p3p_gao` [Gao PAMI 2003], `absolute_pose::upnp` [Kneip ECCV 2014]. **Absolute pose non-minimal solvers documented**: `absolute_pose::epnp` [Lepetit IJCV 2009 — same algorithm as OpenCV's SOLVEPNP_EPNP], `absolute_pose::upnp` (also valid for non-minimal). **Generalized/multi-camera absolute pose solvers documented** via Source #85 §"Non-central absolute pose": `absolute_pose::gp3p` (Kneip 3-point generalized), `absolute_pose::gpnp` [Kneip 2014]. **Non-linear LM optimizer documented**: `absolute_pose::optimize_nonlinear(adapter)` — handles both central + non-central cases; canonical refinement after RANSAC. **RANSAC documented**: `sac::Ransac` + `sac_problems::absolute_pose::AbsolutePoseSacProblem(adapter, algorithm)` with **algorithm parameter selectable from {KNEIP, GAO, EPNP, GP3P}** — richer minimal-solver selection than OpenCV's effectively-4-valid SolvePnPMethod enum (EPNP/AP3P/IPPE/SQPNP after 2 BROKEN entries removed). **CRITICAL input-contract finding**: OpenGV uses **bearing vectors (3D unit vectors)** as input, NOT 2D pixel coordinates — adapters (`AbsoluteAdapterBase`, `RelativeAdapterBase`, `PointCloudAdapterBase`) convert from user data format to OpenGV bearing-vector representation; project must implement adapter or use `CentralAbsoluteAdapter(bearingVectors, points)` constructor where bearingVectors are pre-computed unit vectors via inverse camera-intrinsic projection from C3's pixel correspondences. **CRITICAL threshold-structure finding**: RANSAC threshold is a **3D angle (radians)** between bearing vectors, NOT a 2D pixel reprojection error — Source #85 documents the conversion `ransac.threshold_ = 1.0 - cos(atan(sqrt(2.0)*0.5/800.0))` for a focal length of 800 px and 0.5*sqrt(2.0) pixel reprojection-error-equivalent
+- **Target Audience**: System architects + C4 implementer + Step-7.5 reviewer + license-posture decision-maker (D-C1-1 — BSD-3-Clause-equivalent contingent on Plan-phase license-clearance verification due to NOASSERTION SPDX-detector status) + C7 (Jetson runtime) implementer (canonical OpenGV requires custom build on JetPack 6 ARM Cortex-A78AE — no canonical Jetson distribution; Plan-phase MVE prerequisite)
+- **Research Boundary Match**: **Partial match** for the project's pinned C4 mode (per-frame pose-from-correspondences via classical RANSAC-PnP with paired LM refinement) — algorithm coverage is RICHER than OpenCV at the minimal-solver axis (UPnP for both minimal+non-minimal, GP3P for generalized cameras, 2 P3P variants [Kneip + Gao] vs OpenCV's 1 P3P variant [Ke & Roumeliotis 2017 AP3P]) BUT the input contract (bearing vectors, not pixels) + threshold contract (3D angle, not pixels) + maintenance status (~3 years stale) require Plan-phase mitigation work. **N/A for the project's domain caveat** — OpenGV is a classical algorithm library with no training data; D-C2-1 retrain decision is irrelevant for OpenGV
+- **Summary**: OpenGV is the canonical reference for non-OpenCV PnP minimal solvers + generalized-camera support. **CRITICAL LICENSE FINDING**: License.txt content matches BSD-3-Clause boilerplate (three numbered redistribution conditions including non-endorsement clause) — eligible on every D-C1-1 license-posture choice CONTINGENT on Plan-phase license-clearance verification gate (because GitHub API SPDX detector reports `NOASSERTION`, indicating the License.txt file uses non-standard boilerplate that didn't match the OSI BSD-3-Clause template detection — recommend Plan-phase counsel-review of the License.txt text to confirm BSD-3-Clause-equivalent dual-use compatibility). **CRITICAL MAINTENANCE FINDING**: ~3 years stale at access time (last pushed 2023-06-07; Doxygen documentation portal generated 2018-01-08); ShanghaiTech Mobile Perception Lab's claimed maintenance is contradicted by commit history. **POSITIVE structural findings**: (i) richer minimal-solver coverage than OpenCV (UPnP minimal+non-minimal, GP3P generalized, 2 P3P variants); (ii) canonical reference for non-trivial PnP variants every modern paper compares against; (iii) generalized-camera support (multi-camera rig, non-central absolute pose) — not directly applicable to project's pinned 1× ADTi 20MP nav frame but architecturally cleaner if the project later adds a side-looking camera. **NEGATIVE structural findings**: (iv) bearing-vector input contract requires adapter or pre-computed unit-vector conversion from pixel correspondences (additional engineering vs OpenCV's direct pixel input); (v) 3D-angle RANSAC threshold requires conversion from project's pixel-reprojection-error budget; (vi) NO direct 6×6 covariance output from `optimize_nonlinear` (same finding as OpenCV — D-C4-2 covariance-recovery-strategy applies identically to OpenGV)
+- **Related Sub-question**: SQ3+SQ4 / C4 — OpenGV per-mode API capability verification (Mandatory `context7` lookup NOT-INDEXED + WebFetch fallback PASS per Per-Mode rule item 2; cross-validated against canonical GitHub API metadata WebFetch + canonical License.txt WebFetch + canonical Doxygen documentation portal [Source #85]); **D-C1-1 license-posture compliance**: BSD-3-Clause-equivalent CONTINGENT on Plan-phase license-clearance verification gate (NOASSERTION SPDX-detector caveat); **D-C4-1 carry-forward REINFORCED** (bearing-vector input contract still requires 2D→3D lift on satellite-tile-side from pixel correspondences); **D-C4-2 NEW gate APPLIES IDENTICALLY** to OpenGV (`optimize_nonlinear` returns no covariance — same Plan-phase mitigation strategies as OpenCV); **D-C4-3 NEW gate raised by OpenGV closure** — license-clearance verification due to NOASSERTION SPDX status; **D-C4-4 NEW gate raised by OpenGV closure** — maintenance-staleness mitigation (Plan-phase decision: accept-as-is + freeze upstream / fork into project-controlled branch + apply Eigen-3.4+ + JetPack-6 patches in-house / migrate to Ceres-only as fallback if patches not feasible)
+
+
+### Source #85
+- **Title**: OpenGV canonical Doxygen documentation portal — `laurentkneip.github.io/opengv/page_how_to_use.html` (How to use OpenGV: vocabulary, library organization, adapter pattern interface, conventions, problem types and examples) + `namespaceopengv.html` (top-level namespace) + `namespaceopengv_1_1absolute__pose.html` (absolute-pose methods reference) + `namespaceopengv_1_1relative__pose.html` (relative-pose methods reference) + `namespaceopengv_1_1sac.html` + `namespaceopengv_1_1sac__problems_1_1absolute__pose.html`
+- **Link**: documentation portal entry https://laurentkneip.github.io/opengv/ (accessed 2026-05-08); how-to-use page https://laurentkneip.github.io/opengv/page_how_to_use.html (accessed 2026-05-08; **Doxygen-generated 2018-01-08 21:43:04 by Doxygen 1.8.11 = 8.3 years old at access time**)
+- **Tier**: L1 (canonical project-official Doxygen-generated documentation; the canonical reference for OpenGV's adapter pattern, function signatures, RANSAC integration, and threshold-structure conventions)
+- **Publication Date**: page-generation 2018-01-08; access date 2026-05-08
+- **Timeliness Status**: ⚠️ Established-baseline-reference window with **8.3-year-old documentation** — Plan-phase architect MUST cross-check actual `master` branch source (`opengv/include/opengv/absolute_pose/methods.hpp` + `opengv/include/opengv/sac/Ransac.hpp` + `opengv/include/opengv/sac_problems/absolute_pose/AbsolutePoseSacProblem.hpp`) for any signature drift between 2018 documentation and 2023-06-07 master branch HEAD. The documentation portal is structurally complete for the canonical 2013-2018 published API surface; subsequent commits (2018-2023) appear to be primarily fix commits + ShanghaiTech-era additions
+- **Version Info**: master branch at git commit ea7c66f5e (last commit 2023-06-07). **Pinned-mode runnable example query (2/3) — WebFetch PASS**: Source #85 §"Central absolute pose" provides the canonical OpenGV runnable example: `absolute_pose::CentralAbsoluteAdapter adapter(bearingVectors, points); std::shared_ptr<sac_problems::absolute_pose::AbsolutePoseSacProblem> absposeproblem_ptr(new sac_problems::absolute_pose::AbsolutePoseSacProblem(adapter, sac_problems::absolute_pose::AbsolutePoseSacProblem::KNEIP)); sac::Ransac<sac_problems::absolute_pose::AbsolutePoseSacProblem> ransac; ransac.sac_model_ = absposeproblem_ptr; ransac.threshold_ = 1.0 - cos(atan(sqrt(2.0)*0.5/800.0)); ransac.max_iterations_ = maxIterations; ransac.computeModel(); ransac.model_coefficients_;` followed by optional `absolute_pose::optimize_nonlinear(adapter)` LM refinement on the inlier set with `adapter.sett(initial_translation); adapter.setR(initial_rotation);`. **Disqualifier-probe query (3/3) — FOUR FINDINGS (1 negative-but-mitigable structural + 3 caveats)**: (i) **CRITICAL contract finding — OpenGV uses bearing vectors (3D unit vectors) as input, NOT 2D pixel coordinates** (Source #85 explicit "OpenGV assumes to be in the calibrated case, and landmark measurements are always given in form of bearing vectors in a camera frame"); the project must implement a `CentralAbsoluteAdapter` constructor or pre-compute unit-vector conversion from C3's pixel correspondences via inverse camera-intrinsic projection — additional engineering vs OpenCV's direct pixel input contract; this is an API-level structural difference, not a fundamental algorithmic limitation; (ii) **CRITICAL covariance finding — `optimize_nonlinear` does NOT directly emit a 6×6 pose covariance** (Source #85 documentation does not document a covariance output API; D-C4-2 covariance-recovery-strategy applies identically to OpenGV — Plan-phase mitigation strategies (a) post-hoc Jacobian-based via custom Jacobian propagation through `optimize_nonlinear` residuals OR (b) wrap OpenGV result in GTSAM `Marginals` posterior OR (c) heuristic scaling = AC-NEW-4 REJECT family); (iii) **CRITICAL threshold-structure finding — RANSAC threshold is a 3D angle (radians) between bearing vectors, NOT a 2D pixel reprojection error** (Source #85 §"Ransac threshold" canonical conversion `ransac.threshold_ = 1.0 - cos(atan(sqrt(2.0)*0.5/800.0))` for focal length 800 px and reprojection-error-equivalent 0.5*sqrt(2.0) pixels); project must convert from pixel-reprojection-error budget at runtime; (iv) **CRITICAL maintenance staleness — Doxygen portal generated 2018-01-08 + last commit 2023-06-07 = ~8.3 years documentation staleness + ~3 years code staleness** at access time 2026-05-08; D-C4-4 NEW Plan-phase mitigation strategy required; (v) **License-clearance contingency** — License.txt is BSD-3-Clause-equivalent but GitHub SPDX detector reports NOASSERTION; D-C4-3 NEW Plan-phase license-clearance verification gate required for dual-use deployment compliance
+- **Target Audience**: System architects + C4 implementer + Step-7.5 reviewer + license-posture decision-maker (D-C1-1 + D-C4-3 NEW) + Plan-phase architect (richer-minimal-solver-coverage role documentation for engine Component Option Breadth rule compliance + bearing-vector adapter engineering work + 3D-angle threshold conversion engineering work + D-C4-4 NEW maintenance-staleness mitigation gate)
+- **Research Boundary Match**: Documents the OpenGV library's complete absolute-pose API surface (4 minimal solvers + 2 non-minimal solvers + 1 LM optimizer + 1 RANSAC integration + 4 algorithm-selectable RANSAC enum values) at the structural detail required for Plan-phase decision-making; runnable examples for both central + non-central + relative + multi-camera cases. **N/A for the project's domain caveat** — same as Source #84
+- **Summary**: Canonical Doxygen documentation portal for OpenGV's adapter-pattern interface and method signatures. Documents richer minimal-solver coverage than OpenCV (UPnP for both minimal+non-minimal, GP3P for generalized cameras, 2 P3P variants [Kneip + Gao] vs OpenCV's 1 [AP3P Ke & Roumeliotis 2017]). **CRITICAL contract differences vs OpenCV**: (i) bearing-vector input (3D unit vectors) instead of 2D pixels — adapter required; (ii) 3D-angle RANSAC threshold instead of pixel reprojection — conversion required; (iii) `optimize_nonlinear` LM refinement does not emit covariance — D-C4-2 still applies. **Documentation staleness**: page generated 2018-01-08 by Doxygen 1.8.11 (8.3 years old). **Maintenance staleness**: master branch last pushed 2023-06-07 (~3 years stale). **Recommended pinned mode**: `CentralAbsoluteAdapter` + `AbsolutePoseSacProblem::KNEIP` (Kneip's P3P inside RANSAC) + `optimize_nonlinear` LM refinement — Kneip's P3P is the canonical OpenGV-distinctive minimal solver and is the closest structural analog to OpenCV's `flags=SOLVEPNP_AP3P` (both are P3P variants but Kneip's is the original 2011 method while AP3P is Ke & Roumeliotis 2017 algebraic alternate); for project's planar-scene D-C4-1 = 4-DoF flat-earth lift case, OpenGV does NOT have a dedicated planar-scene minimal solver equivalent to OpenCV's `flags=SOLVEPNP_IPPE` — project would need to use Kneip's P3P or EPNP without the planar-scene specialization advantage. For project's 6-DoF DSM-lift case, OpenGV's UPnP is the modern globally-optimal alternate (analogous structural role to OpenCV's `flags=SOLVEPNP_SQPNP`)
+- **Related Sub-question**: SQ3+SQ4 / C4 — OpenGV per-mode API capability verification (cross-source verification with Source #84 GitHub API + License.txt; runnable example documented; structural caveats documented including bearing-vector contract + 3D-angle threshold + LM-no-covariance findings); **D-C4-2 NEW gate APPLIES IDENTICALLY**; **D-C4-3 NEW gate raised** (license-clearance contingency); **D-C4-4 NEW gate raised** (maintenance-staleness mitigation)
+
+
+### Source #86
+- **Title**: GTSAM canonical implementation — `borglab/gtsam` (Georgia Tech Smoothing and Mapping library; C++ classes for smoothing and mapping in robotics and vision using factor graphs and Bayes networks) GitHub repository metadata via GitHub API + LICENSE + LICENSE.BSD — **BSD-3-Clause** (LICENSE.BSD file contains 3 numbered redistribution conditions including non-endorsement clause; **GitHub API license SPDX detector reports `license.spdx_id: "NOASSERTION"`** because the wrapper LICENSE file at the repo root references `LICENSE.BSD` indirectly + bundles third-party license declarations rather than directly containing OSI canonical BSD-3-Clause boilerplate text; verified BSD-3-Clause via direct WebFetch of `https://raw.githubusercontent.com/borglab/gtsam/develop/LICENSE.BSD`); 3424 stars + 927 forks + 60 subscribers + 140 open issues; created 2017-03-27; **last pushed 2026-05-08T13:00:22Z = TODAY at access time** (daily-active maintenance — fresher than OpenCV); default branch `develop`; size 109374 KB; topics include `estimation, perception, robotics, sensorfusion`; canonical website https://gtsam.org and Doxygen portal https://borglab.github.io/gtsam/. **Bundled third-party libraries** (per LICENSE wrapper file): CCOLAMD 2.9.6 (BSD-3, gtsam/3rdparty/CCOLAMD), Ceres auto-diff/jet code only (BSD-3, modified, gtsam/3rdparty), Eigen 3.3.7 (MPL2 file-level copyleft, gtsam/3rdparty/Eigen), METIS 5.1.0 (Apache-2.0, gtsam/3rdparty/metis), Spectra v0.9.0 (MPL2, externally referenced) — **all clean for project's dual-use deployment** (MPL2 is file-level copyleft only, doesn't propagate to project product code; Apache-2.0 + BSD-3 are permissive)
+- **Link**: GitHub API metadata https://api.github.com/repos/borglab/gtsam (accessed 2026-05-08); canonical repo https://github.com/borglab/gtsam ; LICENSE wrapper https://raw.githubusercontent.com/borglab/gtsam/develop/LICENSE (top-level documents bundled-library licensing); LICENSE.BSD https://raw.githubusercontent.com/borglab/gtsam/develop/LICENSE.BSD (BSD-3-Clause canonical boilerplate "Copyright (c) 2010, Georgia Tech Research Corporation, Atlanta, Georgia 30332-0415, All Rights Reserved" with three numbered redistribution conditions); canonical website https://gtsam.org ; Doxygen portal https://borglab.github.io/gtsam/
+- **Tier**: L1 (project-official codebase by Georgia Tech Research Corporation Borg Lab; canonical reference factor-graph SLAM library used by every modern multi-frame state-estimation deployment as the de-facto industry-standard factor-graph foundation; cited by every C-row component's deployment guide; canonical `LevenbergMarquardtOptimizer` + `Marginals` posterior is the **industry-standard reference for covariance-honest pose estimation**)
+- **Publication Date**: original GTSAM C++ library 2010 (Frank Dellaert + Borg Lab Georgia Tech) → open-source release 2010-12 → migration to GitHub 2017-03-27 → version 4.3a1 indexed in context7 at access time (next-major-release rolling-development branch `develop`); access date 2026-05-08; daily commits to `develop` branch
+- **Timeliness Status**: ✅ Within Established-baseline-reference window (2010+ — established competitive ground for factor-graph SLAM + covariance-honest pose estimation; Established-competitive-mandatory-baseline exemption applies — `LevenbergMarquardtOptimizer` + `Marginals` is the **canonical covariance-honest factor-graph reference** for the C4 row's modern-competitive-lead role and **directly addresses AC-NEW-4 covariance-honesty contract** without D-C4-2 mitigation work)
+- **Version Info**: 4.3a1 at access time (default branch `develop` = next-major-release rolling-development branch; current stable release 4.2 from 2024). **`LevenbergMarquardtOptimizer` + `Marginals` posterior covariance recovery API surface** — see Source #87 below for full documentation and runnable examples
+- **Target Audience**: System architects + C4 implementer + Step-7.5 reviewer + license-posture decision-maker (D-C1-1 — BSD-3-Clause; bundled deps clean) + C5 (state estimator) implementer (GTSAM iSAM2 + factor-graph fusion is the canonical incremental-multi-frame-fusion pathway that scales naturally from C4 single-frame PnP to C5 multi-frame state estimation) + Plan-phase architect (D-C4-2 option (b) Plan-phase pathway candidate)
+- **Research Boundary Match**: **Full match** for the project's pinned C4 mode (per-frame pose-from-correspondences contract on Jetson Orin Nano Super) AT THE COVARIANCE-HONESTY AXIS — GTSAM is the **only C4 candidate evaluated to date that emits 6×6 pose covariance NATIVELY via `Marginals(graph, result).marginalCovariance(pose_key)`** without custom Jacobian engineering. **Architectural extension match**: GTSAM's factor-graph paradigm extends naturally from C4 single-frame PnP to C5 multi-frame state estimation via iSAM2 + `BetweenFactor<Pose3>` + `PriorFactorPose3` — would simplify C5 implementation if both C4 and C5 are GTSAM-based. **N/A for the project's domain caveat** — GTSAM is a classical factor-graph library with no training data; D-C2-1 retrain decision is irrelevant for GTSAM
+- **Summary**: GTSAM is the canonical industry-standard factor-graph SLAM library by Georgia Tech Borg Lab (Frank Dellaert et al.); the `gtsam::slam` module ships `GenericProjectionFactor<Pose3, Point3, CALIBRATION>` as the canonical per-correspondence projection factor for PnP-class problems. **CRITICAL POSITIVE LICENSE FINDING**: BSD-3-Clause via LICENSE.BSD (`Copyright (c) 2010, Georgia Tech Research Corporation`) — permissive, BSD/permissive license track on the C4 modern-competitive-lead axis; **deployment-ready under every D-C1-1 license-posture choice** with the cleanest license-compliance story tied with cvg/LightGlue + DISK + XFeat + OpenCV; bundled dependencies are clean (BSD-3/Apache-2.0/MPL2 file-level — all dual-use compatible). **Daily-active maintenance**: last pushed 2026-05-08 (TODAY at access time) — among the most actively-maintained C-row references; **fresher than OpenCV's last-pushed 2026-05-08T07:00:03Z by 6 hours at access time**. **CRITICAL POSITIVE COVARIANCE FINDING**: `Marginals(graph, result).marginalCovariance(pose_key)` emits a **direct 6×6 pose covariance** with no custom engineering — **the only C4 candidate evaluated to date that satisfies AC-NEW-4 covariance-honesty contract NATIVELY without D-C4-2 mitigation work**; this is the canonical Plan-phase pathway for D-C4-2 = (b) wrap-OpenCV-result-in-GTSAM-Marginals OR full-GTSAM-as-primary
+- **Related Sub-question**: SQ3+SQ4 / C4 — GTSAM per-mode API capability verification (Mandatory `context7` lookup INDEXED at `/borglab/gtsam` with **1121 code snippets at version 4.3a1** — best context7 indexing of any C4 candidate evaluated; full per-mode API documentation accessible via `query-docs` tool); **D-C1-1 license-posture compliance**: BSD-3-Clause with clean bundled deps; **D-C4-2 NATIVELY SATISFIED** via `Marginals` posterior covariance recovery — GTSAM is the canonical Plan-phase pathway for D-C4-2 = (b) wrap-OpenCV-result-in-GTSAM-Marginals OR full-GTSAM-as-primary; **NO new D-C4-N gates raised** by GTSAM closure (D-C4-1 carry-forward applies identically, D-C4-2 natively satisfied)
+
+
+### Source #87
+- **Title**: GTSAM canonical Python documentation via context7-indexed library `/borglab/gtsam` at version 4.3a1 (1121 code snippets) — `python/gtsam/examples/CameraResectioning.ipynb` (canonical PnP example with `LevenbergMarquardtOptimizer`) + `gtsam/slam/doc/ProjectionFactor.ipynb` (`GenericProjectionFactorCal3_S2` API documentation) + `python/gtsam/examples/Pose2SLAMExample.ipynb` + `python/gtsam/examples/PlanarSLAMExample.ipynb` (`Marginals.marginalCovariance` posterior covariance recovery) + `gtsam/inference/doc/FactorGraph.ipynb` (`NonlinearFactorGraph` API documentation)
+- **Link**: context7 library ID `/borglab/gtsam` at version 4.3a1; canonical docs portal https://borglab.github.io/gtsam/ ; canonical Python examples directory https://github.com/borglab/gtsam/tree/develop/python/gtsam/examples (accessed 2026-05-08 via context7 query-docs MCP integration); CameraResectioning canonical example https://github.com/borglab/gtsam/blob/develop/python/gtsam/examples/CameraResectioning.ipynb ; ProjectionFactor canonical documentation https://github.com/borglab/gtsam/blob/develop/gtsam/slam/doc/ProjectionFactor.ipynb
+- **Tier**: L1 (canonical project-official documentation via context7-indexed library; the canonical reference for GTSAM's `GenericProjectionFactorCal3_S2`, `LevenbergMarquardtOptimizer`, `Marginals.marginalCovariance`, `NonlinearFactorGraph`, `Cal3_S2` calibration, `Pose3` 6-DoF pose, and `noiseModel.Diagonal.Sigmas` API surface)
+- **Publication Date**: rolling Jupyter notebook documentation auto-updated on every push to `develop` branch; access date 2026-05-08; canonical PnP example `CameraResectioning.ipynb` has been part of the GTSAM Python distribution since version 4.0 (~2019); access via context7 query at version 4.3a1
+- **Timeliness Status**: ✅ Within Established-baseline-reference window (rolling Jupyter notebook documentation; the canonical reference for GTSAM's PnP + covariance API surface at the project's evaluation time)
+- **Version Info**: 4.3a1 at access time (default branch `develop`). **Mode-enumeration query (1/3) — context7 INDEXED PASS**: `context7 resolve-library-id` returned `/borglab/gtsam` at version 4.3a1 with 1121 code snippets + High source reputation. **Pinned-mode runnable example query (2/3) — context7 query-docs PASS**: canonical PnP runnable Python example from `CameraResectioning.ipynb`: `calibration = Cal3_S2(1, 1, 0, 50, 50)` → `graph = NonlinearFactorGraph()` → per-correspondence factor add via `graph.add(resectioning_factor(measurement_noise, X(1), calibration, Point2(image_pixel), Point3(world_landmark)))` for each 2D-3D correspondence → `initial = Values(); initial.insert(X(1), Pose3(Rot3(...), Point3(...)))` → `result = LevenbergMarquardtOptimizer(graph, initial).optimize()`. **`GenericProjectionFactorCal3_S2` canonical API**: `GenericProjectionFactorCal3_S2(measured_pt2: Point2, pixel_noise: gtsam.noiseModel, pose_key: Symbol, landmark_key: Symbol, calibration: Cal3_S2, body_P_sensor: Pose3=identity)` — per-correspondence projection factor with optional sensor-body offset for IMU-camera extrinsic. **CRITICAL POSITIVE 6×6 covariance recovery API**: `marginals = gtsam.Marginals(graph, result); pose_covariance = marginals.marginalCovariance(pose_key)` — direct 6×6 posterior covariance with NO custom Jacobian engineering required; this is the **DIRECT AC-NEW-4 covariance-honesty contract satisfaction pathway** that no other C4 candidate evaluated to date provides natively. **Disqualifier-probe query (3/3) — TWO FINDINGS (1 negative-but-mitigable structural + 1 caveat)**: (i) **CRITICAL contract finding — GTSAM has NO native RANSAC algorithm** — canonical pattern is to run RANSAC externally (e.g., via OpenCV `cv::solvePnPRansac` for the inlier mask) THEN build the factor graph from inliers only with `GenericProjectionFactorCal3_S2`; alternative is in-graph robust outlier rejection via `gtsam.noiseModel.Robust.Create(gtsam.noiseModel.mEstimator.Huber.Create(1.0), gaussian_noise)` (Huber/Tukey/Cauchy M-estimator robust kernels) OR `GncOptimizer` (Graduated Non-Convexity, Yang et al. RAL 2020) for globally-convergent RANSAC alternative; this couples C4 = GTSAM-as-primary with C5 = OpenCV-RANSAC-as-inlier-detector OR full-GTSAM-with-robust-noise-model OR full-GTSAM-with-GncOptimizer; (ii) **Memory + binary-size CAVEAT — GTSAM library footprint is ~50-200 MB at runtime depending on factor-graph size and bundled-dependency build configuration** (vs OpenCV's ~10-50 MB calib3d module); on Jetson Orin Nano Super 8 GB shared memory budget, GTSAM is the **heaviest C4 candidate evaluated to date** but still well within AC-4.2 budget when co-resident with C1/C2/C3/C5/C6
+- **Target Audience**: System architects + C4 implementer + Step-7.5 reviewer + Plan-phase architect (modern-competitive-lead role documentation for engine Component Option Breadth rule compliance + D-C4-2 NATIVELY SATISFIED + D-C5-N forward-looking carry-forward for state estimator factor-graph extension)
+- **Research Boundary Match**: **Full match** for the C4 row's pinned mode AT THE COVARIANCE-HONESTY AXIS (GTSAM `Marginals.marginalCovariance` is the only C4 candidate evaluated to date that emits 6×6 pose covariance natively; canonical PnP runnable example provided via `CameraResectioning.ipynb`; complete API surface for `LevenbergMarquardtOptimizer` + `GenericProjectionFactorCal3_S2` + `Cal3_S2` + `Pose3` + `noiseModel.Diagonal.Sigmas` documented in canonical Python notebooks); **Architectural-extension match**: GTSAM's factor-graph paradigm extends naturally from C4 single-frame PnP to C5 multi-frame state estimation via iSAM2 + `BetweenFactor<Pose3>` — would simplify C5 implementation if both C4 and C5 are GTSAM-based
+- **Summary**: The canonical GTSAM Python documentation (via context7 at version 4.3a1 with 1121 code snippets) is the definitive reference for `GenericProjectionFactorCal3_S2`, `LevenbergMarquardtOptimizer`, `Marginals.marginalCovariance`, and `NonlinearFactorGraph` API surface. **CRITICAL POSITIVE FINDING for the C4 row**: `Marginals(graph, result).marginalCovariance(pose_key)` emits a **direct 6×6 pose covariance NATIVELY** with no custom Jacobian engineering — **the only C4 candidate evaluated to date that satisfies AC-NEW-4 covariance-honesty contract without D-C4-2 mitigation work**. **NO native RANSAC** — canonical pattern is external RANSAC (via OpenCV solvePnPRansac for inliers) then GTSAM factor-graph from inliers, OR in-graph robust noise model (`gtsam.noiseModel.Robust.Create` + Huber/Tukey/Cauchy), OR `GncOptimizer` (Yang et al. RAL 2020 Graduated Non-Convexity). **Heavier library footprint** than OpenCV (~50-200 MB at runtime) but still well within AC-4.2 8 GB shared memory budget. **Architectural extension to C5**: factor-graph paradigm scales naturally to multi-frame state estimation via iSAM2 + `BetweenFactor<Pose3>` + `PriorFactorPose3` — would simplify C5 implementation
+- **Related Sub-question**: SQ3+SQ4 / C4 — GTSAM per-mode API capability verification (cross-source verification of canonical Python examples + ProjectionFactor API + Marginals posterior + LevenbergMarquardtOptimizer + NonlinearFactorGraph); **D-C4-2 NATIVELY SATISFIED** via `Marginals.marginalCovariance` — GTSAM is the canonical Plan-phase pathway for D-C4-2 = (b); cross-cite to Fact #20 + #21 closures from C2 row (canonical PnP+RANSAC+LM reference pipeline shape feeds AC-NEW-4 covariance-honesty contract); forward-cite to C5 row (factor-graph paradigm extension to multi-frame state estimation via iSAM2)

_docs/00_research/01_source_registry/C5_state_estimator.md

@@ -0,0 +1,95 @@
+# Source Registry — C5: State estimator / sensor fusion
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation). Sources for C5 (state estimator / sensor fusion) candidates.
+>
+> Index: [`00_summary.md`](00_summary.md). Sibling categories: [SQ6](SQ6_external_positioning.md), [SQ1](SQ1_existing_systems.md), [SQ2](SQ2_canonical_pipeline.md), [C1](C1_vio.md), [C2](C2_vpr.md), [C3](C3_matchers.md), [C4](C4_pose_estimation.md). Backing fact cards: [`../02_fact_cards/C5_state_estimator.md`](../02_fact_cards/C5_state_estimator.md). Component fit matrix row: [`../06_component_fit_matrix/C5_state_estimator.md`](../06_component_fit_matrix/C5_state_estimator.md).
+
+---
+
+## Source #88 — Solà 2017 "Quaternion kinematics for the error-state Kalman filter" (canonical aerial/quaternion ESKF tutorial)
+
+**Title**: "Quaternion kinematics for the error-state Kalman filter"
+**Author**: Joan Solà
+**Venue**: arXiv preprint cs.RO 1711.02508 (HAL hal-01122406; Semantic Scholar 12412090e46d1b21eecc59d1326edb8e47e9640e)
+**Submitted**: 2017-11-03 (revision v5 hosted on HAL); originally drafted earlier and continually revised since 2014
+**URL**: <https://arxiv.org/abs/1711.02508> (canonical) + <https://hal.science/hal-01122406v5> (HAL mirror)
+**Tier**: L1 (canonical authoritative tutorial; 592 citations per Semantic Scholar; the de-facto industry reference for ESKF + quaternion algebra in robotics + aerospace + UAV applications since 2017; open-access public-domain academic preprint)
+**Length**: 73 sections including 9 main parts (§1 quaternion definition + §2 rotations + §3 conventions + §4 perturbations/derivatives/integrals + §5 error-state kinematics for IMU-driven systems + §6 fusing IMU with complementary sensory data + §7 ESKF using global angular errors + §8 high-order integration variants + §9 references + §10 appendix)
+**Date Accessed**: 2026-05-08
+
+**Why it matters for C5**:
+- §5.1 lists the THREE structural advantages of ESKF over standard EKF that drive its dominance for UAV applications: (i) minimal orientation error-state (no over-parametrization, no covariance singularity), (ii) error-state always near origin (linearization always valid), (iii) error-state always small (Jacobians fast and often constant).
+- §5.4 provides discrete-time error-state Jacobians directly usable for project's IMU integration.
+- §6 (sub-divided into §6.1 measurement update + §6.2 injection + §6.3 covariance reset) is the canonical recipe for fusing IMU with complementary sensors (project's case = C1 VIO + C4 satellite anchors + FC IMU).
+- §6 explicitly states (line 2013 of the paper text): "At the arrival of other kind of information than IMU, such as GPS or vision, we proceed to correct the ESKF. ... These vision + IMU setups are very interesting for use in **GPS-denied environments**, and can be implemented on mobile devices ... but also on **UAVs and other small, agile platforms**." — a direct project-relevant endorsement from the canonical tutorial.
+- §1675-1677 of the paper text frames the project's exact problem statement: "Integrating IMU readings leads to dead-reckoning positioning systems, which drift with time. Avoiding drift is a matter of fusing this information with absolute position readings such as GPS or vision."
+- §6.3 explicitly notes that the canonical reset Jacobian G can be approximated as `G = I_18` in most implementations, "but the expression here provided should produce more precise results, which might be of interest for reducing long-term error drift in odometry systems" — relevant for project's 8-hour fixed-wing flights where long-term drift is a binding concern.
+- §7 provides an alternate formulation using global angular errors (vs §5's local angular errors); both are valid; project must pick one and stick with it.
+
+---
+
+## Source #89 — Reference open-source ESKF implementations (canonical-paper-derived)
+
+**Repositories examined**:
+
+| # | Repo | Language | License | Sensors fused | Project relevance |
+|---|---|---|---|---|---|
+| 89.a | `ludvigls/ESKF` | Python | (LICENSE not declared in front-page README — Plan-phase verification gate **D-C5-1 NEW** required if adoption) | IMU + GNSS for fixed-wing UAVs | **DIRECTLY MATCHES project hardware family (fixed-wing UAV + IMU + GNSS-replacement)** — closest documentary template; tested on simulated + real datasets per author description |
+| 89.b | `cggos/imu_x_fusion` | C++/ROS | (Plan-phase verification gate **D-C5-1 NEW** required if adoption) | IMU + GNSS + 6DoF-Odom (loosely-coupled) — also IEKF, UKF (UKF/SPKF, JUKF, SVD-UKF), MAP variants | **MATCHES project pattern** — multi-source loosely-coupled fusion (IMU + GNSS-as-satellite_anchor + Odom-as-VIO) |
+| 89.c | `EliaTarasov/ESKF` | C++/ROS | (Plan-phase verification gate **D-C5-1 NEW** required if adoption) | GPS + Magnetometer + Vision Pose + Optical Flow + Range Finder fused with IMU (ROS Error-State Kalman Filter based on PX4/ecl) | **CLOSE MATCH but PX4-derived** — license-clear if PX4/ecl BSD-3-Clause, but verify that the derived code is BSD-3-Clause (PX4 is dual BSD/Apache, ecl is BSD-3-Clause) |
+| 89.d | `koledickarlo/ESKF-ESP32` | C++ | (LICENSE not declared in front-page README — Plan-phase verification gate **D-C5-1 NEW** required if adoption) | Accelerometer + Gyroscope + Optical Flow + Time-of-Flight (microcontroller-class, ESP32) | NOT MATCH — microcontroller-class targets (ESP32) not Jetson; useful only as small-state ESKF reference (Solà 2017 paper explicit citation) |
+| 89.e | `joansola/slamtb` | MATLAB | (LICENSE not declared in front-page README) | EKF-SLAM (full visual-inertial SLAM toolbox) | Author Joan Solà's own SLAM Toolbox in MATLAB — the most authoritative reference for the canonical paper but MATLAB-only, NOT deployable on JetPack 6 |
+
+**Interpretation**: For Fact #88, project does NOT directly reuse any of the above repositories at the source-code level (license verification gates D-C5-1 NEW + cross-domain adaptation costs). Instead, the project implements ESKF following Solà 2017 §5+§6 equations directly in Python (NumPy/SciPy) or C++17 (Eigen3), using ludvigls/ESKF (89.a) as the closest documentary reference template for fixed-wing UAV ESKF structure. The reference implementations serve as evidence that Solà 2017 ESKF is implementable + deployable on UAV-class platforms with multi-sensor fusion patterns identical to the project's pinned configuration.
+
+**URLs accessed (full canonical README pages)**:
+- <https://github.com/ludvigls/ESKF>
+- <https://github.com/cggos/imu_x_fusion>
+- <https://github.com/EliaTarasov/ESKF>
+- <https://github.com/koledickarlo/ESKF-ESP32>
+- <https://github.com/joansola/slamtb>
+
+**Tier**: L1 (canonical project repositories; multiple independent reproductions of Solà 2017 paper across Python, C++/ROS, MATLAB, and microcontroller-class) + L2 (reference template only; project does NOT directly reuse).
+**Date Accessed**: 2026-05-08
+
+---
+
+## Source #90 — GTSAM `ImuFactor` / `CombinedImuFactor` / `PreintegratedImuMeasurements` / `PreintegratedCombinedMeasurements` (context7 query-docs at `/borglab/gtsam` — IMU pre-integration sub-API)
+
+**Title**: GTSAM canonical `ImuFactor` and `CombinedImuFactor` API reference + canonical Python runnable examples
+**Source**: context7 query-docs at `/borglab/gtsam` version 4.3a1 with 1121 code snippets (cross-cite to Source #87 from C4 Fact #54 — same library, different sub-API surface; queried 2026-05-08 for IMU + state-estimation extension to C5)
+**Returned canonical Python notebooks**:
+- `gtsam/navigation/doc/ImuFactor.ipynb` — basic `ImuFactor(X(0), V(0), X(1), V(1), B(0), pim)` 5-key factor + canonical `PreintegrationParams.MakeSharedU(9.81)` setup + `PreintegratedImuMeasurements(params, bias_hat)` + `pim.integrateMeasurement(acc_meas, gyro_meas, dt)` + `pim.predict(initial_state, current_best_bias)` + `imu_factor.evaluateError(pose_i, vel_i, pose_j, vel_j, bias_i)`
+- `gtsam/navigation/doc/CombinedImuFactor.ipynb` — modern `CombinedImuFactor(X(0), V(0), X(1), V(1), B(0), B(1), pim)` 6-key factor with bias evolution per random walk via `PreintegrationCombinedParams.MakeSharedU(9.81)` + `params.setBiasAccCovariance(np.eye(3) * bias_acc_rw_sigma**2)` + `params.setBiasOmegaCovariance(np.eye(3) * bias_gyro_rw_sigma**2)` + `params.setBiasAccOmegaInit(initial_bias_cov)` + `PreintegratedCombinedMeasurements(params, bias_hat)`
+- `gtsam/navigation/doc/PreintegratedImuMeasurements.ipynb` — full PIM workflow: `pim.integrateMeasurement(acc, gyro, dt)` × N → `pim.deltaTij()` / `pim.deltaRij().matrix()` / `pim.deltaPij()` / `pim.deltaVij()` / `pim.biasHat()` / `pim.preintMeasCov()` 9×9 covariance + `pim.predict(initial_state, current_best_bias)` for IMU-only state extrapolation
+- `gtsam/navigation/doc/GPSFactor.ipynb` — `GPSFactor(pose_key, gps_measurement_enu, gps_noise_model)` for 3-DoF GPS prior + `GPSFactorArmCalib(pose_key, lever_arm_key, gps_measurement_enu, gps_noise_model)` for GPS with unknown lever-arm calibration
+
+**Tier**: L1 (canonical context7-indexed library documentation at version 4.3a1; cross-validated against canonical Doxygen portal `borglab.github.io/gtsam/`).
+**URL**: context7 indexing of <https://github.com/borglab/gtsam/tree/develop/gtsam/navigation/doc/> (canonical Borg Lab navigation documentation; access via context7 server at queried-date 2026-05-08)
+**Cross-cite**: Source #86 (canonical `borglab/gtsam` GitHub repo + LICENSE.BSD direct WebFetch — BSD-3-Clause throughout per C4 Fact #54), Source #87 (canonical GTSAM Python examples via context7 query-docs at version 4.3a1 — `CameraResectioning.ipynb` + `Pose2SLAMExample.ipynb` + `PlanarSLAMExample.ipynb` per C4 Fact #54)
+
+**Date Accessed**: 2026-05-08 (~13:00 UTC, immediately after C4 Fact #54 closure — same daily-active GTSAM master branch state)
+
+---
+
+## Source #91 — GTSAM `ISAM2` / `IncrementalFixedLagSmoother` / `Marginals` with iSAM2 results (context7 query-docs at `/borglab/gtsam` — incremental smoothing sub-API)
+
+**Title**: GTSAM canonical `ISAM2` and `IncrementalFixedLagSmoother` incremental smoothing API + `Marginals` posterior recovery for iSAM2 results
+**Source**: context7 query-docs at `/borglab/gtsam` version 4.3a1 with 1121 code snippets (queried 2026-05-08 for incremental smoothing sub-API)
+**Returned canonical Python notebooks**:
+- `gtsam/inference/doc/ISAM.ipynb` — `GaussianISAM(initial_bayes_tree)` constructor + `isam.update(new_factors)` incremental graph modification + `isam.print()` introspection (legacy linear `GaussianISAM`; modern nonlinear `ISAM2` follows the same API pattern with additional `ISAM2Params(relinearizeThreshold, relinearizeSkip, factorization, evaluateNonlinearError, cacheLinearizedFactors, ...)` configuration)
+- `python/gtsam/examples/PlanarSLAMExample.ipynb` — `Marginals(graph, result).marginalCovariance(key)` 6×6 posterior covariance recovery (works with both batch `LevenbergMarquardtOptimizer` results and `ISAM2.calculateEstimate()` results)
+- `python/gtsam/examples/Pose2SLAMExample.ipynb` — same canonical `PriorFactorPose2(1, Pose2(0, 0, 0), PRIOR_NOISE)` initial-pose anchor pattern; reusable for Pose3 (`PriorFactorPose3(X(0), Pose3(...), prior_noise)`) for project's 3D state estimation
+- `gtsam/slam/doc/lago.ipynb` — `lago.initialize(graph)` linear-and-iterative-pose-graph initialization (good for cold-start pose initialization from FC GPS-extrapolated pose at boot per AC-NEW-1)
+- `gtsam/slam/doc/InitializePose3.ipynb` — `InitializePose3.initialize(graph)` chordal-relaxation 3D initialization (modern alternative for Pose3 cold-start)
+- `gtsam/inference/doc/FactorGraph.ipynb` — `NonlinearFactorGraph()` + `BetweenFactorPose2(X(0), X(1), Pose2(1, 0, 0), odometry_noise)` + `PriorFactorPose2(X(0), Pose2(0, 0, 0), prior_noise)` core factor-graph patterns (project applies Pose3 variants: `BetweenFactorPose3` + `PriorFactorPose3` + `GenericProjectionFactorCal3DS2`)
+
+**Note on `IncrementalFixedLagSmoother`**: context7 query-docs at /borglab/gtsam returned ISAM (legacy GaussianISAM) examples but did NOT return a top-3 `IncrementalFixedLagSmoother` snippet on the queried search. The IncrementalFixedLagSmoother class is documented in the canonical GTSAM source tree at `gtsam_unstable/nonlinear/IncrementalFixedLagSmoother.h` (not in the `develop` branch's stable area; in the `gtsam_unstable` namespace, requiring user to opt-in to unstable APIs). Project must verify at Plan-phase Jetson MVE that IncrementalFixedLagSmoother is the correct sliding-window primitive vs writing custom marginalization on top of `ISAM2.marginalizeLeaves(keys_to_marginalize)`.
+
+**Tier**: L1 (canonical context7-indexed library documentation at version 4.3a1) + L2 (IncrementalFixedLagSmoother — gtsam_unstable namespace, verification at Plan phase required).
+**URL**: context7 indexing of <https://github.com/borglab/gtsam/tree/develop/gtsam/inference/doc/> + <https://github.com/borglab/gtsam/tree/develop/python/gtsam/examples/> (canonical Borg Lab inference + examples documentation; access via context7 server at queried-date 2026-05-08)
+**Cross-cite**: Source #86 + Source #87 + Source #90 (all GTSAM library; same daily-active master branch state)
+
+**Date Accessed**: 2026-05-08
+
+---

_docs/00_research/01_source_registry/C6_tile_cache_spatial_index.md

@@ -0,0 +1,142 @@
+# Source Registry — C6: Tile cache + spatial index
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation). Sources backing the C6 component candidates ([`../06_component_fit_matrix/C6_tile_cache_spatial_index.md`](../06_component_fit_matrix/C6_tile_cache_spatial_index.md)) and C6 fact cards ([`../02_fact_cards/C6_tile_cache_spatial_index.md`](../02_fact_cards/C6_tile_cache_spatial_index.md)).
+>
+> Index: [`00_summary.md`](00_summary.md). Sibling component sources: [C1 VIO](C1_vio.md), [C2 VPR](C2_vpr.md), [C3 Matchers](C3_matchers.md), [C4 Pose](C4_pose_estimation.md), [C5 State estimator](C5_state_estimator.md). Sub-question sources: [SQ6 external positioning](SQ6_external_positioning.md), [SQ1 existing systems](SQ1_existing_systems.md), [SQ2 canonical pipeline](SQ2_canonical_pipeline.md).
+
+---
+
+## Scope summary
+
+C6 candidates evaluated documentary level: **Cand 1 (mandatory simple-baseline)** mirrors the parent-suite `satellite-provider` pattern (PostgreSQL + pure btree composite on slippy-map `(tile_zoom, tile_x, tile_y, version)` + filesystem tile storage at `./tiles/{zoom}/{x}/{y}.jpg`); **Cand 2 (modern-competitive-lead-spatial-extension)** = PostGIS GiST on `geography(POINT,4326)` for geographic side + pgvector HNSW for descriptor ANN side + same filesystem tile storage. Both candidates share the same Postgres-as-runtime-DB substrate per user-pinned scope (Postgres on Jetson at runtime, c6_postgres_locus = A). The user explicitly stated the satellite-provider pattern is NOT carved in stone — Cand 2 may cascade changes back to the satellite-provider IF research reveals a MATERIAL improvement (small improvements stay with Cand 1).
+
+---
+
+## Sources
+
+### Source #92 — Parent-suite `satellite-provider` existing pattern (verified directly via filesystem read at /Users/obezdienie001/dev/azaion/suite/satellite-provider/)
+
+**Title**: `azaion/suite/satellite-provider` .NET 8.0 microservice (PostgreSQL + Dapper + filesystem tile storage)
+**Tier**: L1 — primary code in the same multi-repo project workspace
+**URL**: file:///Users/obezdienie001/dev/azaion/suite/satellite-provider/
+**Access date**: 2026-05-08
+**Direct verification**:
+- README at `satellite-provider/README.md` — confirms PostgreSQL backend, .NET 8.0 microservice, Dapper-based DataAccess layer, filesystem tile storage at `./tiles/{zoomLevel}/{x}/{y}.jpg`, NO PostGIS extension declared.
+- Migration `001_CreateTilesTable.sql` — `tiles` table with `(id UUID PK, zoom_level INT, latitude DOUBLE PRECISION, longitude DOUBLE PRECISION, tile_size_meters DOUBLE PRECISION, tile_size_pixels INT, image_type VARCHAR(10), maps_version VARCHAR(50), file_path VARCHAR(500), created_at, updated_at)`.
+- Migration `003_CreateIndexes.sql` — `CREATE INDEX idx_tiles_composite ON tiles(latitude, longitude, tile_size_meters)` + `CREATE INDEX idx_tiles_zoom ON tiles(zoom_level)` + `CREATE INDEX idx_regions_status ON regions(status)`. **Pure btree composite indexes; NO GiST, NO PostGIS, NO spatial extension.**
+- Migration `011_AddTileCoordinates.sql` — RENAME `zoom_level` → `tile_zoom`; ADD `tile_x INT NOT NULL` + `tile_y INT NOT NULL` derived via slippy-map Web Mercator math (`tile_x = FLOOR((longitude + 180.0) / 360.0 * POWER(2, tile_zoom))::INT` + `tile_y = FLOOR((1.0 - LN(TAN(RADIANS(latitude)) + 1.0 / COS(RADIANS(latitude))) / PI()) / 2.0 * POWER(2, tile_zoom))::INT`); CREATE UNIQUE INDEX `idx_tiles_unique_location ON tiles(latitude, longitude, tile_zoom, tile_size_meters, version)` + `CREATE INDEX idx_tiles_coordinates ON tiles(tile_zoom, tile_x, tile_y, version)`. **Confirms: existing pattern uses btree on slippy-map (zoom, x, y) integer-coordinate columns for spatial-grid range queries.**
+
+**Key facts extracted**:
+- DB engine: PostgreSQL (vanilla, no extensions).
+- Spatial index strategy: pure btree composite on slippy-map integer coordinates `(tile_zoom, tile_x, tile_y, version)` for spatial-grid range queries; secondary btree on lat/lon for inverse-geocode lookups.
+- Tile bytes: filesystem at canonical slippy-map path `./tiles/{zoom}/{x}/{y}.jpg`.
+- DB ↔ filesystem coupling: `file_path VARCHAR(500)` pointer in DB.
+- Migration mechanism: numbered SQL files as `EmbeddedResource`, run automatically on startup via `DatabaseMigrator.cs`.
+- App layer: .NET 8.0 + Dapper + raw SQL repos.
+
+**Implication**: For the on-Jetson C6 (which is Python/C++, not .NET), the equivalent stack is `psycopg[binary]` or `asyncpg` Python driver + raw SQL queries against the same schema pattern.
+
+---
+
+### Source #93 — PostgreSQL official documentation: btree multi-column index ordering and range query optimization
+
+**Title**: PostgreSQL 16 documentation — "Multicolumn Indexes" + "Indexes and ORDER BY" + "EXPLAIN" + "btree access method"
+**Tier**: L1 — official authoritative docs
+**URL**: <https://www.postgresql.org/docs/current/indexes-multicolumn.html> + <https://www.postgresql.org/docs/current/btree.html>
+**Access date**: 2026-05-08
+**Direct verification**: pending WebFetch
+**Key facts to extract**:
+- Btree multicolumn index supports range queries on the leading prefix (i.e., `WHERE tile_zoom = ? AND tile_x BETWEEN ? AND ?` uses the index optimally).
+- Btree composite index access time: O(log N) where N = total rows.
+- Storage overhead: typically ~50-100 bytes per index entry depending on column types.
+
+**Use**: backs Fact #92 sub-matrix entries on AC-4.1 (latency) and AC-4.2 (memory) for Cand 1.
+
+---
+
+### Source #94 — PostGIS official documentation: GiST spatial index on geography type + KNN distance ordering
+
+**Title**: PostGIS 3.4 documentation — "GiST Indexes" + "geography Type" + "PostGIS Special Functions Index" + "ST_DWithin" + "<-> KNN operator"
+**Tier**: L1 — official authoritative docs (OGC SFS-compliant canonical extension)
+**URL**: <https://postgis.net/docs/using_postgis_dbmanagement.html#idx-spgist> + <https://postgis.net/docs/geography.html> + <https://postgis.net/workshops/postgis-intro/knn.html>
+**Access date**: 2026-05-08
+**Direct verification**: pending WebFetch
+**Key facts to extract**:
+- GiST index access time on `geography(POINT,4326)`: O(log N) for bounding-box pre-filter; full geographic distance check is exact (not approximate).
+- KNN ordering via `ORDER BY position <-> ST_MakePoint(?, ?)::geography LIMIT K` is index-optimized in PostGIS 2.0+.
+- `ST_DWithin(position::geography, ST_MakePoint(?, ?)::geography, radius_m)` supports radius queries with native great-circle distance.
+- PostGIS extension installed footprint: typically ~30-50 MB shared libraries + ~10-20 MB SRID/projection metadata catalog.
+
+**Use**: backs Fact #93 sub-matrix entries on AC-4.1 (latency) and AC-4.2 (memory) for Cand 2 + comparative-improvement-vs-Cand-1 analysis.
+
+---
+
+### Source #95 — pgvector official documentation: HNSW index for vector similarity search
+
+**Title**: pgvector — "Open-source vector similarity search for Postgres" (`pgvector/pgvector`)
+**Tier**: L1 — canonical implementation by Andrew Kane
+**URL**: <https://github.com/pgvector/pgvector> + context7 indexed via `/pgvector/pgvector`
+**Access date**: 2026-05-08
+**Direct verification**: pending context7 + WebFetch
+**Key facts to extract**:
+- HNSW index API: `CREATE INDEX ON items USING hnsw (embedding vector_l2_ops)` + `CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops)` + `CREATE INDEX ON items USING hnsw (embedding vector_ip_ops)`.
+- Default tunable parameters: `m=16` (max connections per layer) + `ef_construction=64` (build-time candidate list size); query-time `ef_search` (default 40).
+- Vector dimension limits: pgvector 0.7+ supports up to 16,000 dimensions for HNSW; 2,000 dimensions for IVFFlat.
+- Memory footprint: extension itself ~5-10 MB shared library; per-vector storage = 4 bytes × dimensions (so 2048-D = 8 KB/vec, 1024-D = 4 KB/vec, 512-D = 2 KB/vec, 256-D = 1 KB/vec).
+
+**Use**: backs Fact #93 sub-matrix on descriptor ANN side for Cand 2 + comparative cache footprint analysis.
+
+---
+
+### Source #96 — FAISS official documentation: in-memory ANN library + Python bindings
+
+**Title**: FAISS — "A library for efficient similarity search and clustering of dense vectors" (`facebookresearch/faiss`)
+**Tier**: L1 — canonical implementation by Meta AI Research
+**URL**: <https://github.com/facebookresearch/faiss> + <https://faiss.ai/>
+**Access date**: 2026-05-08
+**Direct verification**: pending WebFetch + context7
+**Key facts to extract**:
+- Index types relevant to C6 descriptor ANN: `IndexFlatL2` (brute-force, exact), `IndexHNSWFlat` (HNSW graph, approximate), `IndexIVFFlat` (Inverted File, approximate w/ training).
+- Memory: in-memory only at query time; loaded from disk via `faiss.read_index(path)` at startup.
+- License: MIT.
+- Python API: `faiss.IndexFlatL2(d)` / `faiss.IndexHNSWFlat(d, m)` / `index.add(xb)` / `D, I = index.search(xq, k)`.
+
+**Use**: backs Fact #92 sub-matrix on descriptor ANN side for Cand 1 (app-side FAISS in-memory loaded at takeoff from Postgres bytea blobs).
+
+---
+
+### Source #97 — Postgres on NVIDIA Jetson Orin Nano memory footprint and JetPack 6 deployment
+
+**Title**: PostgreSQL on ARM64 / Ubuntu 22.04 (JetPack 6 base) — official packaging + Docker images
+**Tier**: L1 — official Postgres ARM64 packages + Docker `postgres:16-alpine` image documentation
+**URL**: <https://hub.docker.com/_/postgres> + <https://www.postgresql.org/download/linux/ubuntu/>
+**Access date**: 2026-05-08
+**Direct verification**: pending WebFetch
+**Key facts to extract**:
+- ARM64 packages available for Postgres 16 on Ubuntu 22.04 (JetPack 6 base).
+- Default `shared_buffers=128MB` + `work_mem=4MB` resident footprint ~80-150 MB on idle; ~200-400 MB under modest load.
+- Docker `postgres:16-alpine` image size: ~250 MB compressed.
+- PostGIS Docker image `postgis/postgis:16-3.4-alpine` adds ~50-80 MB to base postgres image.
+
+**Use**: backs both Fact #92 + Fact #93 sub-matrix entries on AC-4.2 (8 GB shared memory budget) for the Postgres-on-Jetson deployment.
+
+---
+
+### Source #98 — Slippy Map Tilenames specification (OpenStreetMap canonical reference)
+
+**Title**: Slippy Map Tilenames — XYZ tile coordinate system + Web Mercator projection
+**Tier**: L1 — canonical convention documented by OpenStreetMap Foundation
+**URL**: <https://wiki.openstreetmap.org/wiki/Slippy_map_tilenames>
+**Access date**: 2026-05-08
+**Direct verification**: pending WebFetch
+**Key facts to extract**:
+- Tile X/Y math: `xtile = floor((lon + 180) / 360 * 2^zoom)` + `ytile = floor((1 - asinh(tan(lat * π/180)) / π) / 2 * 2^zoom)` — matches satellite-provider migration 011 exactly.
+- Tile coverage: at zoom Z, world divided into 2^Z × 2^Z tiles; each tile covers `360/2^Z` longitude × variable-latitude.
+- Project zoom: ZoomLevel 18 (per satellite-provider README default) covers ~38m × 38m at equator (cited as "tileSizeMeters: 38.2" in README sample response).
+- Cache budget per AC-8.3 (10 GB): at typical JPEG ~30 KB/tile, fits ~330,000 tiles = roughly an area of 50 km × 50 km × 9 zoom levels OR a single mission corridor at zoom 18 of ~1000 km × 12 m.
+
+**Use**: backs both Fact #92 + Fact #93 sub-matrix entries on AC-8.3 (10 GB cache budget) + AC-3.x (mission corridor coverage).
+
+---
+
+(Subsequent sources #99+ added during fact extraction below as candidate-specific evidence is gathered.)

_docs/00_research/01_source_registry/C7_inference_runtime.md

@@ -0,0 +1,190 @@
+# Source Registry — C7: On-Jetson inference runtime
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation). Sources backing the C7 cross-cutting integration row ([`../06_component_fit_matrix/C7_inference_runtime.md`](../06_component_fit_matrix/C7_inference_runtime.md)) and C7 fact cards ([`../02_fact_cards/C7_inference_runtime.md`](../02_fact_cards/C7_inference_runtime.md)).
+>
+> Index: [`00_summary.md`](00_summary.md). Sibling component sources: [C1 VIO](C1_vio.md), [C2 VPR](C2_vpr.md), [C3 Matchers](C3_matchers.md), [C4 Pose](C4_pose_estimation.md), [C5 State estimator](C5_state_estimator.md), [C6 Tile cache](C6_tile_cache_spatial_index.md). Sub-question sources: [SQ6 external positioning](SQ6_external_positioning.md), [SQ1 existing systems](SQ1_existing_systems.md), [SQ2 canonical pipeline](SQ2_canonical_pipeline.md).
+
+---
+
+## Scope summary
+
+C7 is a **cross-cutting integration row** rather than a per-component candidate row: it pins how the C1 VIO learned-frontend (if any), C2 VPR backbone, and C3 matcher actually run on the Jetson Orin Nano Super under JetPack 6 — TensorRT vs ONNX Runtime+TRT EP vs pure PyTorch FP16. Per the user-pinned scope (locked via `/autodev` AskQuestion 2026-05-08 — see `_docs/_autodev_state.md` `c7_breadth=B`, `c7_quantization=A`, `c7_overkill_options=A`), three documentary candidate rows are evaluated: **TensorRT native primary** + **ONNX Runtime + TensorRT EP interop alternate** + **pure PyTorch FP16 mandatory simple-baseline**. INT8 primary + FP16 fallback per candidate; INT8-only candidates Experimental until calibration data exists. Triton / DeepStream / CUDA-Python custom kernels noted-and-rejected in one sentence (server/video-pipeline class or out-of-budget for embedded 8 h mission). Cand-row candidates inherit and propagate Plan-phase gates already opened by C2 (D-C2-5 DINOv2 ViT-export to TensorRT FP16/INT8) and C3 (D-C3-2 LightGlue inference runtime path).
+
+---
+
+## Sources
+
+### Source #99 — NVIDIA TensorRT 10.x official documentation portal (context7-indexed)
+
+**Title**: NVIDIA TensorRT — SDK for optimizing and accelerating deep learning inference on NVIDIA GPUs (mixed precision, dynamic shapes, transformer optimizations)
+**Tier**: L1 — official authoritative SDK documentation (NVIDIA primary)
+**URL**: <https://docs.nvidia.com/deeplearning/tensorrt/latest/> + context7 indexed at `/websites/nvidia_deeplearning_tensorrt`
+**Access date**: 2026-05-08
+**Direct verification**: ✅ context7 query "INT8 calibration EntropyCalibrator2 ICudaEngine deserialize Jetson Orin Nano FP16 mixed precision deployment workflow Python builder" returned 9371 code snippets at Source Reputation High + Benchmark Score 75.25.
+
+**Key APIs verified**:
+- **INT8 calibrator hierarchy**: `nvinfer1::IInt8Calibrator` (abstract base) + `nvinfer1::IInt8EntropyCalibrator` (deprecated) + `nvinfer1::IInt8EntropyCalibrator2` (current canonical) + `nvinfer1::IInt8MinMaxCalibrator`. Each defines `getBatchSize()` + `getBatch(void* bindings[], const char* names[], int32_t nbBindings)` + `readCalibrationCache(size_t& length)` + `writeCalibrationCache(const void* ptr, size_t length)` + `getAlgorithm()` returning `kENTROPY_CALIBRATION_2` for the canonical path.
+- **Python builder INT8 enable pattern** (canonical TensorRT 10.x):
+  ```python
+  config.set_flag(trt.BuilderFlag.INT8)
+  config.int8_calibrator = Int8_calibrator
+  Int8_calibrator = EntropyCalibrator(["input_node_name"], batchstream)
+  ```
+- **Mixed-precision flag pattern**: `config.set_flag(trt.BuilderFlag.FP16)` + `config.set_flag(trt.BuilderFlag.INT8)` for combined FP16+INT8 mixed precision (TensorRT auto-selects per-layer precision based on calibration data).
+
+**Use**: backs Fact #94 (TensorRT native primary candidate) per-mode API verification block + Plan-phase D-C7-1 calibration-dataset-strategy + D-C7-2 mixed-precision flag matrix.
+
+---
+
+### Source #100 — Microsoft ONNX Runtime official documentation (context7-indexed) + Jetson AI Lab community wheel index
+
+**Title**: Microsoft ONNX Runtime — cross-platform ML inference and training accelerator with TensorRT execution provider; Jetson-specific install path via Jetson AI Lab community PyPI index
+**Tier**: L1 — official authoritative SDK documentation (Microsoft primary) + L2 community-maintained Jetson wheel index
+**URL**: <https://onnxruntime.ai/> + context7 indexed at `/microsoft/onnxruntime` (v1.25.0) + <https://pypi.jetson-ai-lab.io/jp6/cu126/> + <https://github.com/dusty-nv/jetson-containers/issues/1283> + <https://github.com/microsoft/onnxruntime/issues/20503> + <https://github.com/microsoft/onnxruntime/issues/27562>
+**Access date**: 2026-05-08
+**Direct verification**: ✅ context7 query "TensorRT execution provider TrtFp16Enable TrtInt8Enable TrtCachePath onnxruntime-gpu Jetson ARM64 inference session options" returned 1462 code snippets at Source Reputation High + Benchmark Score 82.23 (highest of the 3 C7 candidate context7 lookups).
+
+**Key APIs verified**:
+- **Provider enumeration + config pattern** (canonical Python API):
+  ```python
+  import onnxruntime as ort
+  print(ort.get_available_providers())
+  tensorrt_options = {'device_id': 0, 'trt_max_workspace_size': 2147483648, 'trt_fp16_enable': True}
+  cuda_options = {'device_id': 0, 'arena_extend_strategy': 'kNextPowerOfTwo', 'gpu_mem_limit': 2 * 1024 * 1024 * 1024}
+  session_trt = ort.InferenceSession(
+      "model.onnx",
+      providers=[('TensorrtExecutionProvider', tensorrt_options), ('CUDAExecutionProvider', cuda_options), 'CPUExecutionProvider']
+  )
+  ```
+- **Provider-cascade behavior**: ORT TRT EP attempts to optimize each subgraph via TensorRT; falls back to CUDA EP for unsupported ops; falls back to CPU EP if neither GPU EP applies. Subgraph fallback is automatic and per-op transparent.
+
+**Jetson install constraints (CRITICAL)**:
+- **Standard `pip install onnxruntime-gpu` does NOT work on Jetson Tegra** — Microsoft does not publish prebuilt aarch64 wheels with CUDA/TensorRT EPs (per Issue #20503: "NVIDIA does not have CI infrastructure to publish them").
+- **Canonical install path (JetPack 6 + CUDA 12.6 + Ubuntu 22.04)**: `pip3 install onnxruntime-gpu --index-url https://pypi.jetson-ai-lab.io/jp6/cu126`.
+- **Alternate index (CUDA 12.9 + Ubuntu 24.04)**: `pip3 install onnxruntime-gpu --index-url https://pypi.jetson-ai-lab.io/jp6/cu129`.
+- **Known incompatibility**: onnxruntime-gpu v1.23.0 wheels for JetPack 6 were built against `numpy<2.0.0`; importing under `numpy>=2.0.0` raises a compatibility error per Issue #27562. Pin numpy<2 in project requirements until upstream rebuild is published.
+- **Standard pip install `onnxruntime` (CPU-only) succeeds but exposes only `CPUExecutionProvider` and `AzureExecutionProvider`** — does NOT include CUDA EP or TensorRT EP.
+
+**Use**: backs Fact #95 (ONNX Runtime + TensorRT EP interop alternate candidate) per-mode API verification block + Plan-phase D-C7-3 ORT-Jetson-wheel-pin + D-C7-4 numpy-version-pin.
+
+---
+
+### Source #101 — PyTorch official documentation (context7-indexed) + Jetson AI Lab PyTorch wheel availability for JetPack 6
+
+**Title**: PyTorch — open-source machine learning framework (tensor computation with strong GPU acceleration; tape-based autograd); Jetson-specific wheels available via Jetson AI Lab + NVIDIA forums
+**Tier**: L1 — official authoritative SDK documentation (PyTorch Foundation primary) + L1 NVIDIA Developer Forums (canonical Jetson PyTorch distribution channel)
+**URL**: <https://pytorch.org/docs/stable/amp.html> + context7 indexed at `/pytorch/pytorch` (v2.5.1, v2.8.0, v2.9.1, v2.11.0) + <https://forums.developer.nvidia.com/t/installing-pytorch-for-jetpack-6-2/349519> + <https://forums.developer.nvidia.com/t/jetpack-6-2-and-pytorch-2-6-0-on-jetson-nano-orin/331972>
+**Access date**: 2026-05-08
+**Direct verification**: ✅ context7 query "torch.cuda.amp.autocast half precision FP16 inference mode no_grad CUDA Jetson Orin ARM64 model.half() torch.compile inference deployment" returned 4866 code snippets at Source Reputation High + Benchmark Score 76.69.
+
+**Key APIs verified**:
+- **`torch.amp.autocast(device_type, dtype, enabled, cache_enabled)`** — canonical AMP context manager (since PyTorch 1.10). Replaces deprecated `torch.cuda.amp.autocast`. Inference pattern:
+  ```python
+  with torch.no_grad():
+      with torch.autocast(device_type='cuda', dtype=torch.float16, enabled=True):
+          output = model(input)
+  ```
+- **`torch.compile(model, backend='inductor')`** — graph-mode optimization for further speedup; tradeoff is cold-start compile cost (~10-60 sec depending on model complexity).
+- **`model.half()`** — eager-mode FP16 weight conversion (full-precision FP16 throughout, vs autocast's per-op precision selection).
+
+**Jetson install constraints**:
+- **Standard `pip install torch` does NOT include CUDA support on Jetson** — must use NVIDIA-published or Jetson AI Lab community wheels.
+- **JetPack 6.2 + CUDA 12.6 + Ubuntu 22.04 + Python 3.10 canonical wheel**: `torch-2.9.0-cp310-cp310-linux_aarch64.whl` from Jetson AI Lab (per NVIDIA forum recommendation). Earlier stable combination: PyTorch 2.5 + torchvision 0.20.
+- **Known dependency issues**: missing `libcudss.so.0` and `libnvdla_runtime.so` on PyTorch 2.9 cu129 wheel under JetPack 6.2 (CUDA 12.6) — version mismatch between wheel build target and installed JetPack CUDA. Mitigation: prefer the cu126 variant for JetPack 6.2.
+- **CUDA capability**: Jetson Orin Nano Super GPU = compute capability **SM 87** (Ampere class).
+
+**Use**: backs Fact #96 (pure PyTorch FP16 mandatory simple-baseline candidate) per-mode API verification block + D-C7-5 PyTorch-Jetson-wheel-pin.
+
+---
+
+### Source #102 — Ultralytics YOLO26 benchmark suite on Jetson Orin Nano Super (April 2026)
+
+**Title**: Update NVIDIA Jetson Orin Nano Super benchmarks with YOLO26 (Ultralytics 8.4.33; commit 8d4e6e8 April 2026)
+**Tier**: L1 — official authoritative benchmark suite (Ultralytics is the canonical YOLO maintainer)
+**URL**: <https://github.com/ultralytics/ultralytics/pull/24097> + <https://github.com/ultralytics/ultralytics/commit/8d4e6e841c89f6598b322695cb2bc816eeba8b93>
+**Access date**: 2026-05-08
+**Direct verification**: ✅ Web search results explicitly cite the per-export-format inference times measured on Jetson Orin Nano Super.
+
+**Key data extracted (YOLO26n on Jetson Orin Nano Super, April 2026 measurement)**:
+
+| Export format | Inference time (ms) | mAP50-95 | Speedup vs FP32 | Accuracy delta vs FP16 |
+|---|---|---|---|---|
+| TensorRT FP32 | 7.53 | 0.4770 | 1.00× | — |
+| TensorRT FP16 | 4.57 | 0.4800 | 1.65× | baseline (slightly higher than FP32 due to noise) |
+| TensorRT INT8 | 3.80 | 0.4490 | 1.98× | **-6.5% mAP50-95** |
+
+**Key data extracted (YOLOv8s on Jetson Orin Nano, NVIDIA forum)**:
+- **INT8**: ~157 QPS (~6.4 ms/inference)
+- **FP16**: ~103 QPS (~9.7 ms/inference)
+- **INT8 vs FP16 speedup**: ~1.5× (vs ~1.20× on YOLO26n — model architecture and memory bandwidth dependent)
+
+**Use**: backs Fact #94 (TensorRT) latency claims for object-detection-class CNN backbones on Jetson Orin Nano Super; provides empirical anchor for the engine's "INT8 primary + FP16 fallback" precision strategy. Caveat: YOLO is a detection network; feature-matching networks (LightGlue / DISK / XFeat) are known to be more quantization-sensitive (see Source #103).
+
+---
+
+### Source #103 — LightGlue ONNX Runtime + TensorRT acceleration (canonical reference) + FP8 ModelOpt quantization findings (Fabio Sim's Journal)
+
+**Title**: Accelerating LightGlue Inference with ONNX Runtime and TensorRT (Fabio Sim's Journal, canonical author of `fabio-sim/LightGlue-ONNX`) + FP8 Quantized LightGlue in TensorRT with NVIDIA Model Optimizer (subsequent post)
+**Tier**: L1 — canonical author of the canonical LightGlue ONNX/TensorRT export pathway (already cited as Source #73 in C3 row)
+**URL**: <https://fabio-sim.github.io/blog/accelerating-lightglue-inference-onnx-runtime-tensorrt/> + <https://fabio-sim.github.io/blog/fp8-quantized-lightglue-tensorrt-nvidia-model-optimizer/> + <https://github.com/qdLMF/LightGlue-with-FlashAttentionV2-TensorRT> (community Jetson Orin NX TensorRT 8.5.2 + FlashAttentionV2 plugin reference implementation)
+**Access date**: 2026-05-08
+**Direct verification**: ✅ Web search results explicitly cite the 2-4× ONNX Runtime + TensorRT speedup over compiled PyTorch and the FP8 5.97× / 0.32× engine-size results.
+
+**Key data extracted**:
+- **LightGlue (transformer-based feature matcher) — ONNX Runtime + TensorRT inference**: 2-4× speedup over compiled PyTorch across various batch sizes and sequence lengths.
+- **FP8 quantized LightGlue (NVIDIA ModelOpt) on Hopper/Ada/Blackwell**:
+  - Engine size ~0.32× of FP32 (~68% smaller).
+  - Up to 5.97× speedup vs FP32.
+  - **Material accuracy degradation**: "match counts dropped. Sometimes they dropped hard." This is qualitatively different from YOLO-class detection networks where INT8 is well-tolerated.
+  - **FP8 hardware support**: requires Hopper / Ada / Blackwell architecture. **Jetson Orin Nano Super is Ampere (SM 87) — NOT FP8-native**. FP8 ModelOpt path applies only via INT8 emulation fallback on Ampere.
+- **Two FP8 formats**: E4M3 (4 exponent bits + 3 mantissa bits, better precision for activations) + E5M2 (5 exponent bits + 2 mantissa bits, better dynamic range for gradients).
+- **Community Jetson reference implementation**: `qdLMF/LightGlue-with-FlashAttentionV2-TensorRT` deploys on Jetson Orin NX 8 GB with TensorRT 8.5.2 + custom FlashAttentionV2 plugin.
+
+**Use**: backs Fact #94 (TensorRT) feature-matching-network INT8 caveat; backs the "INT8-only candidates Experimental until calibration data exists" engine ruling per user-pinned `c7_quantization=A` scope; raises Plan-phase gate D-C7-6 INT8-vs-FP16-per-model-family-precision-policy.
+
+---
+
+### Source #104 — JetPack SDK release notes (NVIDIA official) — JetPack 6.0 / 6.1 / 6.2 version matrix
+
+**Title**: NVIDIA JetPack 6.x SDK Release Notes — TensorRT/CUDA/cuDNN versions per release; Super Mode introduction in JetPack 6.2 (January 2025)
+**Tier**: L1 — official authoritative release notes (NVIDIA Developer)
+**URL**: <https://developer.nvidia.com/embedded/jetpack-sdk-60> + <https://developer.nvidia.com/embedded/jetpack-sdk-61> + <https://developer.nvidia.com/embedded/jetpack-sdk-62> + <https://developer.nvidia.com/blog/nvidia-jetpack-6-2-brings-super-mode-to-nvidia-jetson-orin-nano-and-jetson-orin-nx-modules/>
+**Access date**: 2026-05-08
+**Direct verification**: ✅ Web search results explicitly enumerate TensorRT / CUDA / cuDNN per JetPack release.
+
+**Key data extracted**:
+
+| JetPack | CUDA | TensorRT | cuDNN | Super Mode | Released |
+|---|---|---|---|---|---|
+| 6.0 | 12.2 | 8.6 | 8.9 | No | early 2024 |
+| 6.1 | 12.6 | 10.3 | 9.3 | MAXN mode (dev kit only) | mid-2024 |
+| **6.2** | **12.6** | **10.3** | **9.3** | **YES — Orin Nano Super + Orin NX production modules** | **2025-01-16** |
+
+- **Super Mode performance gains** (vs base Orin Nano): up to 2× higher generative AI inference performance, 70% AI TOPS increase, 50% memory bandwidth boost.
+- **TensorRT 10.3** is the canonical inference runtime version for JetPack 6.1 / 6.2 deployments. Major API upgrade from TensorRT 8.x → 10.x — `IInt8EntropyCalibrator2` API surface is preserved; `INetworkDefinition` and `IBuilderConfig` semantics unchanged.
+
+**Use**: pins the project's target software stack to **JetPack 6.2 + CUDA 12.6 + TensorRT 10.3 + cuDNN 9.3 + Super Mode enabled** for the Jetson Orin Nano Super target hardware. Backs Facts #94, #95, #96 deployability claims.
+
+---
+
+### Source #105 — TensorRT-on-Jetson canonical install constraints (Ultralytics issue reports + NVIDIA forum)
+
+**Title**: TensorRT 10.x on Jetson Orin Nano — install path, hardware-specificity, memory-pressure-during-build constraints
+**Tier**: L2 — community-reported issues with NVIDIA-acknowledged root causes (high signal-to-noise on canonical constraints)
+**URL**: <https://github.com/ultralytics/ultralytics/issues/18882> ("TensorRT does not currently build wheels for Tegra systems") + <https://forums.developer.nvidia.com/t/tensorrt-10-7-0-on-orin-nano/364236> (SM 87 compute-capability mismatch) + <https://github.com/ultralytics/ultralytics/issues/18730> (laptop-GPU-built engine cannot load on Jetson) + <https://github.com/ultralytics/ultralytics/issues/21281> (TensorRT export memory pressure on Orin AGX)
+**Access date**: 2026-05-08
+**Direct verification**: ✅ Web search returned direct issue links with NVIDIA-confirmed root causes.
+
+**Key constraints extracted** (CRITICAL for C7 deployment design):
+
+1. **TensorRT Python wheels are NOT installed via pip on Jetson Tegra**. Standard `pip install tensorrt` raises: `RuntimeError: TensorRT does not currently build wheels for Tegra systems`. The canonical install path is the JetPack-bundled TensorRT (already present after `apt install nvidia-jetpack`), accessed via the system Python at `/usr/lib/python3.10/dist-packages/tensorrt`.
+2. **TensorRT engines are hardware-specific** — engines built against a laptop / dev-machine GPU CANNOT be loaded on the Jetson at runtime. **Engines must be built directly on the Jetson target**.
+3. **GPU compute capability mismatch is silent at build-time, fatal at load-time**: laptop GPUs (e.g., RTX 4090 = SM 89) and Jetson Orin Nano Super (SM 87) produce incompatible engines; the build emits no error, the load logs `Target GPU SM 87 is not supported by this TensorRT release` — version-and-SM-compatibility matrix must be respected.
+4. **TensorRT engine builds on Jetson under memory pressure can segfault during tactic profiling** (8 GB shared CPU+GPU is tight; a rich layer-fusion search consumes peak RAM during `tactic.profile` phase). Mitigation: limit `config.max_workspace_size` to a fraction of the budget (e.g., 1-2 GB) and avoid concurrent inference / Postgres / FAISS during builds.
+5. **JetPack 6.x ships the canonical TensorRT version** (TensorRT 10.3 for JP 6.1/6.2 per Source #104); upgrading TensorRT independently of JetPack is not officially supported.
+
+**Use**: drives D-C7-7 build-on-Jetson-vs-prebuilt-engine-shipping-strategy + D-C7-8 max-workspace-size-cap-for-build-stability + D-C7-9 SM-compatibility-version-pin.
+
+---
+
+(Subsequent sources #106+ added during fact extraction below as candidate-specific evidence is gathered. Closure target: 3 candidate rows + 1 cross-cutting integration matrix.)

_docs/00_research/01_source_registry/C8_fc_adapter.md

@@ -0,0 +1,97 @@
+# Source Registry — C8: MAVLink / MSP2 FC adapter
+
+> Mode A Phase 2 — engine Step 2 (Source Tiering & Exhaustive Web Investigation). C8 batch 1 sources for the FC adapter (per-FC adapter pattern verified at SQ6 closure: ArduPilot Plane via MAVLink `GPS_INPUT`, iNav via `MSP2_SENSOR_GPS` primary OR UBX-impersonation alternate). Confidence labels per `references/source-tiering.md`. Cross-references back to SQ6 fact card sources (#4, #9, #10, #12, #13, #15) where the iNav inbound-handler reality and MSP2/UBX transport options were originally established.
+>
+> Index: [`00_summary.md`](00_summary.md). Sibling component categories: [C1 VIO](C1_vio.md), [C2 VPR](C2_vpr.md), [C3 Matchers](C3_matchers.md), [C4 Pose](C4_pose_estimation.md), [C5 State estimator](C5_state_estimator.md), [C6 Tile cache](C6_tile_cache_spatial_index.md), [C7 Inference runtime](C7_inference_runtime.md). Cross-cuts: [SQ6 external positioning](SQ6_external_positioning.md).
+
+## Sources
+
+### Source #106 — ArduPilot Pymavlink (context7-indexed `/ardupilot/pymavlink`)
+- **Tier**: L1 (canonical Python MAVLink implementation maintained by ArduPilot)
+- **Found via**: context7 `resolve-library-id` → `/ardupilot/pymavlink` → `query-docs` for GPS_INPUT send patterns
+- **Library posture**: 32 code snippets indexed in context7 (Source Reputation: High); coverage emphasizes the JavaScript MAVLink generator output, with thinner Python-side examples in context7 — supplementary primary sources (canonical pymavlink GitHub README + ArduPilot GPS_INPUT dev docs Source #107) carry the canonical Python `master.mav.gps_input_send(...)` send pattern.
+- **License**: LGPL v3 (pymavlink itself); MAVLink generated dialects are MIT — the project's runtime dependency is on the LGPL pymavlink Python package. **Compatible with project's Apache-2.0 dual-use track**: LGPL allows linking from a non-LGPL application without "infecting" application license; the only obligation is to publish/redistribute any modifications to pymavlink itself (project does not modify pymavlink), and to allow users to relink against an updated pymavlink (trivially satisfied for an open-source / company-internal deployment with published `requirements.txt`).
+- **Critical-novelty-sensitivity**: Established baseline; no time window — pymavlink has been the canonical Python MAVLink stack since 2010+, and `GPS_INPUT` (msg 232) has been in `common.xml` since 2017 ArduPilot dev iteration.
+- **Per-mode capability verification (context7 + SQ6 Source #4 AP_GPS_MAV.cpp cross-cite)**: ✅ `GPS_INPUT` decoder confirmed in AP_GPS_MAV.cpp master per SQ6 Fact #1; Python sender uses `master = mavutil.mavlink_connection(...)` + `master.mav.gps_input_send(time_usec, gps_id, ignore_flags, time_week_ms, time_week, fix_type, lat, lon, alt, hdop, vdop, vn, ve, vd, speed_accuracy, horiz_accuracy, vert_accuracy, satellites_visible, yaw)` per pymavlink generated dialect.
+- **Used to support**: Fact #97 (ArduPilot Plane FC adapter primary candidate).
+
+### Source #107 — ArduPilot Plane Non-GPS Position Estimation + MAVProxy GPS Input module documentation
+- **Tier**: L1 (official ArduPilot dev docs portal; documented configuration + canonical injection example)
+- **Found via**: web search for `pymavlink GPS_INPUT msg 232 example ArduPilot Plane non-GPS external positioning companion computer 2025`
+- **Date accessed**: 2026-05-08
+- **URLs**:
+  - https://ardupilot.org/dev/docs/mavlink-nongps-position-estimation.html
+  - https://ardupilot.org/plane/docs/common-non-gps-navigation-landing-page.html
+  - https://ardupilot.org/mavproxy/docs/modules/GPSInput.html
+  - https://ardupilot.org/plane/docs/common-companion-computers.html
+- **Critical configuration captured**: `GPS1_TYPE = 14` (MAVLink) is required on the FC for `GPS_INPUT` ingestion. Without this parameter set, AP_GPS will not accept the message. `EK3_SRC1_POSXY = 3` (GPS) selects the GPS_INPUT-fed virtual GPS as the primary horizontal-position source. Per ArduPilot dev docs, the **preferred method** for non-GPS navigation is `ODOMETRY` or `VISION_POSITION_ESTIMATE` at ≥4 Hz — but `GPS_INPUT` remains supported and is the right choice when the project's outcome contract is "WGS84 coordinates as a real-GPS replacement" (AC-4.3 wording aligns with GPS_INPUT semantics, not ODOMETRY semantics).
+- **Cross-cite**: SQ6 Fact #1 (AP_GPS_MAV.cpp ingestion path) + SQ6 Fact #4 (`ODOMETRY`-velocity-only NOT supported) — together these pin `GPS_INPUT` as the right transport for the project's `{satellite_anchored, visual_propagated, dead_reckoned}` source-label scheme.
+- **Per-mode capability verification**: ✅ All required ACs (AC-4.3 / AC-NEW-2 / AC-NEW-4 / AC-NEW-8) map directly into GPS_INPUT field semantics per SQ6 working summary table.
+
+### Source #108 — pyubx2 (context7-indexed `/semuconsulting/pyubx2` + canonical GitHub README)
+- **Tier**: L1 (canonical Python UBX/NMEA/RTCM3 parser; benchmark score 86.8 in context7; 139 code snippets)
+- **Found via**: context7 `resolve-library-id` → `/semuconsulting/pyubx2` → `query-docs` for UBX-NAV-PVT message construction with full attribute control + serialize-to-bytes pattern for UART transmission
+- **Library posture**: BSD-3-Clause license (clean, dual-use compatible); semuconsulting publishes both the canonical GitHub repo + comprehensive readthedocs.io documentation also indexed in context7 as `/websites/semuconsulting_pyubx2` (239 additional code snippets, benchmark 85.2). The library supports `UBXMessage(ubxClass, ubxID, mode, **kwargs)` constructor with three modes: `GET (0x00)` for output from the receiver, `SET (0x01)` for command input, `POLL (0x02)` for query input. NAV-PVT belongs to the GET output set.
+- **Critical-novelty-sensitivity**: Library/SDK API behaviour — must reflect currently shipped version; semuconsulting/pyubx2 is daily-active (last released 2025).
+- **Per-mode capability verification (context7-confirmed)**: ✅ NAV-PVT message construction with all UBX-NAV-PVT fields supported as keyword arguments per `UBXMessage('NAV', 'NAV-PVT', GET, iTOW=..., year=..., lon=..., lat=..., height=..., hMSL=..., hAcc=..., vAcc=..., velN=..., velE=..., velD=..., gSpeed=..., headMot=..., sAcc=..., headAcc=..., pDOP=..., fixType=..., flags=..., numSV=..., valid=...)`. ✅ `serialize()` method returns the full UBX wire-format bytestring (sync-bytes 0xB5 0x62 + class + ID + length + payload + 8-bit Fletcher checksum). ✅ `parsebitfield=1` mode allows individual bit attributes for `flags` (e.g., `gnssFixOK`, `diffSoln`, `psmState`) and `valid` (e.g., `validDate`, `validTime`, `fullyResolved`, `validMag`) — required for the impersonation path to set the `gnssFixOK` bit that iNav's `gpsMapFixType()` validates.
+- **Used to support**: Fact #98 (iNav UBX impersonation alternate candidate).
+
+### Source #109 — u-blox NEO-M9N Integration Manual (UBX-19014286) + u-blox 8/M8 Receiver Description (UBX-13003221) — UBX-NAV-PVT canonical specification
+- **Tier**: L1 (vendor-authoritative protocol specification PDFs)
+- **Found via**: web search for `UBX-NAV-PVT frame structure spec u-blox protocol M8 M9 fix type fabricate inject iNav 2025`
+- **Date accessed**: 2026-05-08
+- **URLs**:
+  - https://content.u-blox.com/sites/default/files/NEO-M9N_Integrationmanual_UBX-19014286.pdf
+  - https://content.u-blox.com/sites/default/files/products/documents/u-blox8-M8_ReceiverDescrProtSpec_UBX-13003221.pdf
+- **Frame structure captured**: NAV-PVT (class=0x01, ID=0x07) carries 92-byte payload — `iTOW (u32 ms)` + `year (u16)` + `month/day/hour/min/sec (u8 each)` + `valid (u8 bitmask)` + `tAcc (u32 ns)` + `nano (i32 ns)` + `fixType (u8 enum: 0=NoFix, 1=DeadReck, 2=2D, 3=3D, 4=GNSS+DR, 5=TimeOnly)` + `flags (u8 bitmask incl. gnssFixOK bit 0)` + `flags2 (u8)` + `numSV (u8)` + `lon (i32 deg×1e-7)` + `lat (i32 deg×1e-7)` + `height (i32 mm above ellipsoid)` + `hMSL (i32 mm above mean sea level)` + `hAcc (u32 mm)` + `vAcc (u32 mm)` + `velN/velE/velD (i32 each mm/s)` + `gSpeed (i32 mm/s)` + `headMot (i32 deg×1e-5)` + `sAcc (u32 mm/s)` + `headAcc (u32 deg×1e-5)` + `pDOP (u16 ×0.01)` + reserved bytes + `headVeh (i32)` + `magDec (i16)` + `magAcc (u16)`. M9N supersedes M8 with refined NAV-PVT semantics; both are accepted by iNav 9.0 (per Source #11 in SQ6 — UBX ≥ 15.00 protocol version).
+- **Critical-novelty-sensitivity**: Established baseline + library/SDK API behaviour — u-blox NAV-PVT is a stable protocol surface since u-blox 8 (2014); minor field semantics evolve across vendor protocol versions, so exact wire format must be checked against the iNav-target version (iNav 9.0 expects ≥ 15.00).
+- **Per-mode capability verification**: ✅ NAV-PVT contains all fields needed for iNav's `gpsMapFixType()` validation (Source #110 cross-cite): `flags` byte bit 0 `gnssFixOK` + `fixType` enum + `numSV` + `hAcc/vAcc` for AC-NEW-4 covariance honesty.
+- **Used to support**: Fact #98 (iNav UBX impersonation alternate candidate) NAV-PVT frame fabrication spec.
+
+### Source #110 — iNav `gps_ublox.c` source (master, GitHub) — UBX validation gates that the impersonation must pass
+- **Tier**: L1 (canonical iNav firmware source, master branch, accessed via cached web fetch)
+- **Found via**: web search for `iNav GPS UBX validation fixType numSat hDOP threshold reject GNSS spoofing companion computer 2025`
+- **URL**: https://github.com/iNavFlight/inav/blob/master/src/main/io/gps_ublox.c
+- **Date accessed**: 2026-05-08
+- **Critical-novelty-sensitivity**: Library/SDK API behaviour — must reflect current shipped iNav version. iNav 9.0 master (post-2025-12-11 wiki update per SQ6 Source #10) confirmed via direct file read.
+- **Validation logic captured (line-numbered evidence)**:
+  - **Line 215-220**: `gpsMapFixType(fixValid, ubloxFixType)` returns `GPS_FIX_2D` if `fixValid && ubloxFixType == FIX_2D`, returns `GPS_FIX_3D` if `fixValid && ubloxFixType == FIX_3D`, otherwise `GPS_NO_FIX`. **THIS IS THE GATE** the impersonation must pass.
+  - **Line 654**: NAV-PVT path computes `next_fix_type = gpsMapFixType(_buffer.pvt.fix_status & NAV_STATUS_FIX_VALID, _buffer.pvt.fix_type)`. The `fix_status & NAV_STATUS_FIX_VALID` masks the lowest bit of NAV-PVT's `flags` byte (bit 0 = `gnssFixOK`).
+  - **Lines 656-683**: NAV-PVT-driven full state population including `lon (1e-7 deg)`, `lat (1e-7 deg)`, `altitude_msl (mm)`, NED velocity (mm/s converted to cm/s), `speed_2d (mm/s)`, `heading_2d (deg×1e-5 → deg×10)`, `satellites`, `horizontal_accuracy (mm)`, `vertical_accuracy (mm)`, `position_DOP`, valid date/time bits.
+  - **Lines 1024-1060**: Configuration logic — for u-blox version ≥ 15.0 (iNav 9.0+), iNav configures NAV-PVT-only via `configureMSG(MSG_CLASS_UBX, MSG_PVT, 1)`. For older receivers, configures the legacy NAV-POSLLH + NAV-SOL + NAV-VELNED + NAV-TIMEUTC quad. **Implication**: companion impersonator should advertise version ≥ 15.0 via NAV-VER (CLASS=0x0A, ID=0x04) to drive iNav into the simpler NAV-PVT-only protocol.
+- **Per-mode capability verification**: ✅ Validation gate fully decoded; impersonation viability confirmed at the firmware-source level (no opaque downstream filter discovered).
+- **Used to support**: Fact #98 — provides the iNav-firmware-side validation contract that the UBX impersonation must satisfy.
+
+### Source #111 — iNav `docs/development/msp/README.md` (master, GitHub) — MSP2_SENSOR_GPS canonical payload specification
+- **Tier**: L1 (canonical iNav protocol-reference documentation, master branch, accessed via cached web fetch)
+- **Found via**: web search for `MSP2_SENSOR_GPS Python library iNav msp2 protocol companion computer external GPS injection 2025 2026`
+- **URL**: https://github.com/iNavFlight/inav/blob/master/docs/development/msp/README.md
+- **Date accessed**: 2026-05-08
+- **Payload structure captured (line 2999-3031 of the master README)**: `MSP2_SENSOR_GPS (7939 / 0x1F03)` — request payload 36 bytes containing `instance (u8)` + `gpsWeek (u16)` + `msTOW (u32 ms)` + `fixType (u8 = gpsFixType_e)` + `satellitesInView (u8)` + `hPosAccuracy (u16 mm)` + `vPosAccuracy (u16 mm)` + `hVelAccuracy (u16 cm/s)` + `hdop (u16 ×0.01)` + `longitude (i32 deg×1e7)` + `latitude (i32 deg×1e7)` + `mslAltitude (i32 cm)` + `nedVelNorth (i32 cm/s)` + `nedVelEast (i32 cm/s)` + `nedVelDown (i32 cm/s)` + `groundCourse (u16 deg×100)` + `trueYaw (u16 deg×100, 65535 = unavailable)` + `year (u16)` + `month/day/hour/min/sec (u8 each)`. **Reply payload: None.** **Notes: Requires `USE_GPS_PROTO_MSP`. Calls `mspGPSReceiveNewData()`.**
+- **Critical-novelty-sensitivity**: Library/SDK API behaviour — verified against iNav master (post-9.0).
+- **Per-mode capability verification**: ✅ Full payload spec covers all AC-NEW-4 covariance honesty fields (`hPosAccuracy`, `vPosAccuracy`, `hVelAccuracy`); ✅ AC-NEW-8 graceful-degrade signal carried via `fixType` enum (`gpsFixType_e`) — companion can emit `GPS_NO_FIX` (0) or `GPS_FIX_2D` (1) for the "covariance >100 m" / "covariance >500 m" thresholds; ✅ AC-1.4 95% covariance proxy carried in `hPosAccuracy`.
+- **Used to support**: Fact #99 (iNav MSP2_SENSOR_GPS primary candidate).
+
+### Source #112 — Python MSP2 implementations: YAMSPy + INAV-Toolkit `inav_msp.py`
+- **Tier**: L2 (community implementations; NOT vendor-canonical but actively maintained)
+- **Found via**: web search for Python MSP2_SENSOR_GPS libraries; iNav Issue #4465 confirms YAMSPy as community-recommended; agoliveira/INAV-Toolkit confirmed via direct GitHub source read
+- **URLs**:
+  - YAMSPy mention: https://github.com/iNavFlight/inav/issues/4465
+  - INAV-Toolkit `inav_msp.py`: https://github.com/agoliveira/INAV-Toolkit/blob/5c4ef789068399b4dc7461b71c6f71c25aef5e4e/inav_msp.py
+- **Date accessed**: 2026-05-08
+- **Library posture**:
+  - **YAMSPy** (`thecognifly/YAMSPy`): MIT-licensed Python library with explicit MSP V2 support; community-blessed for iNav external-device communication per the iNav issue thread.
+  - **INAV-Toolkit `inav_msp.py`**: 951-line MIT-licensed module implementing `msp_v2_encode(cmd, payload)` + `msp_v2_decode(buffer)` with CRC-8 DVB-S2 checksumming + serial transport. Direct primary-source implementation reference for MSP V2 frame construction.
+- **Critical-novelty-sensitivity**: Library/SDK API behaviour — both libraries are recent (post-2024 commits). **Risk**: community libraries may lag the iNav protocol surface (e.g., MSP V2 sensor message range 0x1F00-0x1FFF was added later than the original MSP V2 baseline). The project may need to either (a) extend the chosen community library with MSP2_SENSOR_GPS-specific encoding helpers, or (b) implement a thin custom encoder using the canonical msp_v2_encode primitive — both paths verified feasible from primary sources.
+- **License notes**: MIT throughout — clean dual-use compatible.
+- **Per-mode capability verification**: ⚠️ MSP V2 frame envelope (0x24 + 'X' + 0x3C + flag + cmd_lo + cmd_hi + len_lo + len_hi + payload + CRC8-DVB-S2) confirmed via INAV-Toolkit primary source; ✅ MSP2_SENSOR_GPS payload structure confirmed via Source #111. Combining the two yields a complete companion-side encoder for the iNav primary path.
+- **Used to support**: Fact #99 (iNav MSP2_SENSOR_GPS primary candidate, Python implementation path).
+
+### Source #113 — iNav `src/main/msp/msp_protocol_v2_sensor.h` (master, GitHub) — MSP2 sensor command-ID range
+- **Tier**: L1 (canonical iNav firmware source, master branch)
+- **Found via**: web search co-result with Source #112; opens via the `msp_protocol_v2_sensor.h` direct link
+- **URL**: https://github.com/iNavFlight/inav/blob/master/src/main/msp/msp_protocol_v2_sensor.h
+- **Date accessed**: 2026-05-08
+- **Critical fact captured**: `MSP2_SENSOR_GPS = 0x1F03` (= 7939 decimal); MSP V2 sensor-message range `0x1F00-0x1FFF` is reserved for sensor injection plugins. iNav 9.0 master expectation: MSP2 frame must use the MSP V2 envelope (sync = 0x24 0x58 0x3C; flag = 0x00; cmd = LE 16-bit; len = LE 16-bit; CRC = CRC-8 DVB-S2 over flag through end of payload).
+- **Per-mode capability verification**: ✅ MSP2_SENSOR_GPS = 0x1F03 confirmed at source; ✅ MSP V2 envelope spec confirmed.
+- **Used to support**: Fact #99 — provides the canonical MSP V2 sensor-message-range definition.