Update autopilot workflow to enhance task management automation

- Adjusted the workflow to automatically revert to New Task after project completion without requiring user confirmation.
- Updated the autopilot state to indicate the current step as `done` and status as `completed`.
- Improved clarity in deployment status reporting by detailing non-deployed services and their functions.

These modifications streamline task management processes and enhance documentation clarity.

.woodpecker/01-test.yml

@@ -0,0 +1,29 @@
+when:
+  event: [push, pull_request, manual]
+  branch: [dev, stage, main]
+
+labels:
+  platform: arm64
+
+steps:
+  - name: unit-tests
+    image: docker:24
+    environment:
+      COMPOSE_PROJECT_NAME: ai-training-tests
+    commands:
+      - mkdir -p tests/test-results
+      - docker compose -f docker-compose.test.yml up --build --abort-on-container-exit --exit-code-from test-runner
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock
+
+  - name: report
+    image: docker:24
+    when:
+      status: [success, failure]
+    environment:
+      COMPOSE_PROJECT_NAME: ai-training-tests
+    commands:
+      - docker compose -f docker-compose.test.yml down -v || true
+      - test -f tests/test-results/test-results.xml && head -40 tests/test-results/test-results.xml || echo "no results xml"
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock