[AZ-808] [AZ-809] [AZ-810] [AZ-811] [AZ-812] Cycle 8 retro + close

Cycle 8 retrospective (cycle-end mode): 5 tickets shipped, 17 SP, 4
batches across 1 cycle theme (strict input validation for the 4
AZ-795 child endpoints + AZ-812 region API field rename).

Artifacts:
- _docs/06_metrics/retro_2026-05-23_cycle8.md
- _docs/06_metrics/structure_2026-05-23_cycle8.md (gap-filled; last
  structural snapshot was cycle 5)

Key cycle-8 findings (now in _docs/LESSONS.md ring buffer):
- Step-14 security-audit Medium findings under the small-fix
  threshold should be resolved in-cycle, not deferred (F-AZ809-1
  closed in commit 8fca6e0, ~30 min from discovery to fix).
- Retro recommendations ship end-to-end when they name concrete
  tickets/files + size as a coherent cycle theme (cycle 7 Action 3
  -> cycle 8 strict-validation slate, first end-to-end traceable
  cross-cycle improvement action in project history).
- Contract wire-format updates (new required field / rename) need a
  ripgrep probe across all consumer paths (perf script, probe
  scripts, README, deploy docs, OpenAPI examples) — partial syncs
  surface at Step 15 perf gate (PT-06 missed AZ-809 requestMaps +
  createTilesZip, fixed in commit 32bc5c1).

Carry-overs to cycle 9: track PT-07 cache-pollution false positive
(harness, not regression), reduce 3 cycles in a row of misleading
"PT-07 fails on warm/cold ratio" entries in the perf report.

Marks Step 17 completed; cycle 8 closed. Next /autodev invocation
starts cycle 9 from Step 0.

Co-authored-by: Cursor <cursoragent@cursor.com>

_docs/LESSONS.md

@@ -37,11 +37,17 @@ If the enum's wire string happens to match a member name case-insensitively (e.g
 
 ## Ring buffer (last 15 entries — newest at top)
 
+- [2026-05-23] [process] Step-14 security-audit Medium findings whose remediation fits the small-fix threshold (≤2 SP, ≤2 files, ≤1 contract bump) should be resolved within the same autodev invocation rather than deferred to cycle N+1 — the fix lands with the same commit chain that introduced the surface, the contract version reflects the fix immediately, and the traceability matrix and blackbox-tests.md sub-cases are written while the finding is fresh; codify the option as a first-class A/B/C choice in `.cursor/skills/autodev/flows/existing-code.md` Step-14 action (cycle 8: F-AZ809-1 unbounded `geofences.polygons` DoS — discovered in commit `ac40a8b`, resolved in commit `8fca6e0` with `MaxPolygons = 50` cap + unit + integration test + `route-creation.md` v1.0.1 patch bump, ~30 minutes from finding to fix landed).
+  Source: _docs/06_metrics/retro_2026-05-23_cycle8.md
+- [2026-05-23] [process] Retrospective recommendations ship end-to-end in the next cycle only when they (a) name concrete tracker tickets / files / endpoints in the action text, (b) are sized as a coherent cycle theme rather than scattered one-off fixes, and (c) the next cycle's planning phase pulls the recommendation directly into the task slate without re-deriving scope — phrase recommendations to satisfy all three or they become multi-cycle carry-overs (cycle 7 Action 3 named the 4 AZ-795 child endpoints + the SP sizing → cycle 8 shipped AZ-808 + AZ-809 + AZ-810 + AZ-811 + AZ-812 as the coherent strict-validation theme, first directly-traceable cross-cycle improvement-action end-to-end in project history).
+  Source: _docs/06_metrics/retro_2026-05-23_cycle8.md
+- [2026-05-23] [tooling] When a contract introduces a new required field or renames a wire key, the test-spec sync step must ripgrep all consumer paths (`scripts/run-performance-tests.sh`, `scripts/probe_*.sh`, `README.md` example URLs, `_docs/04_deploy/*.md` example bodies, `_docs/02_document/tests/blackbox-tests.md` trigger excerpts, OpenAPI examples) and verify they reference the new shape — otherwise partial syncs surface at Step 15 Performance Test gate or worse in production (cycle 8: AZ-812 lat/lon rename updated the perf script but AZ-809's newly-required `requestMaps` + `createTilesZip` fields did not propagate to PT-06; caught at Step 15 with a 1-line script fix in commit `32bc5c1` — a static-check probe at batch closure would have caught it 2 days earlier).
+  Source: _docs/06_metrics/retro_2026-05-23_cycle8.md
 - [2026-05-23] [process] When verifying a "no-regression" AC for an input-validation change ("AZ-NNN does not break existing tests"), the only sound evidence is a green integration-test run — tracing fixture variables back to their generators in source is insufficient because helpers can produce values outside the new bounds and previously slipped through silently when no validator existed; document the standard as "verified by reading source" → unconfirmed, "verified by full test run" → confirmed, and gate the batch report's AC table on the latter before the implement skill closes the batch (cycle 8: AZ-810 batch_04 AC-9 claimed "no AZ-488 regression" based on tracing `latitude = coord.Latitude` in test source, but `NextTestCoordinate` seeded by `(Ticks/TicksPerSecond) % 1_000_000` produced lat far above 90° at runtime; the false-PASS only surfaced at autodev Step 11 when the integration test run returned HTTP 400 from the new validator on the AZ-488 happy path).
   Source: _docs/03_implementation/batch_04_cycle8_report.md (AC-9 row)
-- [2026-05-22] [process] When the implement skill ships a cycle's batch commit without writing `_docs/03_implementation/implementation_report_*_cycle{N}.md`, downstream skills (test-spec cycle-update, document task mode, retrospective Step 1) must fall back to reading the cycle's task specs in `_docs/02_tasks/done/` plus the commit body via `git log --grep='[AZ-...]'` — codify the fallback in those skills' instructions instead of leaving it as per-cycle improvisation, because the implicit contract between Step 10 and Steps 11-17 broke silently this cycle and only succeeded because every downstream skill happened to be robust enough to substitute (cycle 7: AZ-794+AZ-795+AZ-796 shipped as commit `865dfdb` with no report artifact; doc-skill auto-walked the diff, test-spec read the task specs, retrospective wrote from the deploy + security + perf reports — all worked, but the contract was never formal).
+- [2026-05-22] [process] When the implement skill ships a cycle's batch commit without writing `_docs/03_implementation/implementation_report_*_cycle{N}.md`, downstream skills (test-spec cycle-update, document task mode, retrospective Step 1) must fall back to reading the cycle's task specs in `_docs/02_tasks/done/` plus the commit body via `git log --grep='[AZ-...]'` — codify the fallback in those skills' instructions instead of leaving it as per-cycle improvisation, because the implicit contract between Step 10 and Steps 11-17 broke silently this cycle and only succeeded because every downstream skill happened to be robust enough to substitute (cycle 7: AZ-794+AZ-795+AZ-796 shipped as commit `865dfdb` with no report artifact; doc-skill auto-walked the diff, test-spec read the task specs, retrospective wrote from the deploy + security + perf reports — all worked, but the contract was never formal). **Status**: closed by cycle 8 — the implement skill self-corrected and produced both per-batch reports AND a consolidated implementation report AND a completeness report AND a cumulative cross-batch review.
   Source: _docs/06_metrics/retro_2026-05-22_cycle7.md
-- [2026-05-22] [testing] When a strict-validation layer ships (`JsonSerializerOptions.UnmappedMemberHandling.Disallow`, FluentValidation rules, explicit DTO `[JsonRequired]`), expect the project's own integration tests to surface latent bugs the prior lenient defaults had been masking — silent PascalCase fallback property names, out-of-range fixture coordinates, wrong-cased JSON keys; correct them in the same PR or the test suite goes red and the strict layer looks like a regression instead of the bug-finder it is (cycle 7: `IdempotentPostTests.RoutePoint` had been posting `{"Lat":...}` against a `[JsonPropertyName("lat")]` DTO for months; the new strict deserializer caught it and the 2-line payload fix landed alongside the strict layer).
+- [2026-05-22] [testing] When a strict-validation layer ships (`JsonSerializerOptions.UnmappedMemberHandling.Disallow`, FluentValidation rules, explicit DTO `[JsonRequired]`), expect the project's own integration tests to surface latent bugs the prior lenient defaults had been masking — silent PascalCase fallback property names, out-of-range fixture coordinates, wrong-cased JSON keys; correct them in the same PR or the test suite goes red and the strict layer looks like a regression instead of the bug-finder it is (cycle 7: `IdempotentPostTests.RoutePoint` had been posting `{"Lat":...}` against a `[JsonPropertyName("lat")]` DTO for months; the new strict deserializer caught it and the 2-line payload fix landed alongside the strict layer; **cycle 8 instance**: `UavUploadTests.NextTestCoordinate` produced lat > 90°, caught by AZ-810 validator, 2-file clamp fix in batch 4).
   Source: _docs/06_metrics/retro_2026-05-22_cycle7.md
 - [2026-05-22] [architecture] Contract MAJOR bumps for projects with ≤2 known consumers should ship without a wire-format adapter — the cost of maintaining a dual-accepting endpoint outweighs the benefit when the operator runbook can coordinate the consumer cut-over directly; only invest in an adapter when consumer count or release-cadence asymmetry makes coordinated cut-over impractical (cycle 7: `tile-inventory.md` 1.0.0 → 2.0.0 renamed `tileZoom/tileX/tileY → z/x/y` and shipped breaking; the only consumer is `gps-denied-onboard` and the AZ-777 follow-up loop handled the coordination via the operator runbook in `deploy_cycle7.md`).
   Source: _docs/06_metrics/retro_2026-05-22_cycle7.md
@@ -61,11 +67,3 @@ If the enum's wire string happens to match a member name case-insensitively (e.g
   Source: _docs/06_metrics/retro_2026-05-12_cycle4.md
 - [2026-05-12] [process] When a scope-protected task newly *exposes* a pre-existing bug elsewhere in the codebase (vs. introducing a new one), surface it as a recommended follow-up PBI in the batch report AND list it as a "newly exposed bug" separate from "newly introduced findings" in the deploy report — bugs that already existed don't count as cycle-introduced regressions, but they must not be silently re-buried (cycle 4: AZ-500's bootstrap fix unmasked the pre-existing `scripts/run-performance-tests.sh:417` `grep -o | wc -l` + `pipefail` bug).
   Source: _docs/06_metrics/retro_2026-05-12_cycle4.md
-- [2026-05-12] [process] When a cycle has a single non-functional task (migration / refactor / dependency hygiene), the retro must reframe the metric set around continuity (0 regressions), forward-resolution (prior findings closed by the bump itself), and unblocking (capabilities now exercisable end-to-end) — task count + complexity points read as misleading flatlines that look like under-productivity (cycle 4: AZ-500 alone delivered 5 SP vs cycle 3's 18 SP, but the cycle's value was forward-resolving 2 cycle-3 advisories and finally executing PT-01..PT-08 end-to-end against the migrated build).
-  Source: _docs/06_metrics/retro_2026-05-12_cycle4.md
-- [2026-05-12] [process] For cross-team blockers (admin team must supply config values, etc.), prefer an Option-B forcing function (ship the validation/scaffolding with prod-empty config that fails-fast at deploy) over deferring the entire task — the fail-fast contract makes the cross-team conversation impossible to skip and ships the in-workspace work in the current cycle (cycle 3: AZ-494 shipped iss/aud validation with empty prod appsettings so deploy must supply real values).
-  Source: _docs/06_metrics/retro_2026-05-12_cycle3.md
-- [2026-05-12] [process] ACs that prescribe a specific measurement or sentinel mechanism (e.g. "per-item latency < 50ms", "guard fires when DB name contains _test") should also prescribe — or explicitly defer — the path for collecting / enforcing it, or implementations will substitute proxies / equivalents that look like spec drift in review (cycle 3: AZ-492 PT-08 per-item gate cost became a derived proxy; AZ-493 DB-name guard became Host-allowlist).
-  Source: _docs/06_metrics/retro_2026-05-12_cycle3.md
-- [2026-05-12] [process] ACs that require cross-repo writes should be tagged with the target workspace and rendered separately in the traceability matrix — mixing them with in-workspace ACs makes "correctly deferred" indistinguishable from "incomplete work" (cycle 3: AZ-494 AC-7 deferred for the suite-repo write; matrix renders as `◐ deferred` which is ambiguous).
-  Source: _docs/06_metrics/retro_2026-05-12_cycle3.md