Tasks: Integration tests and OpenAPI specs

Phase 3 of 3. DAG of implementation tasks derived from the approved design.

Task list

• [x] 1. Scenario extraction library (the_loop/scenarios)
  • Line-based Gherkin extraction with comment-marker stripping; Scenario dataclass; collect_scenarios over globs; built-in default globs.
  • Depends on: none
  • Requirements: R1, R2
  • Test: pytest cli/tests/test_cli.py -k extract or collect (red→green)
• [x] 2. the-loop scenarios command
  • Register the command; glob resolution (--glob > config > defaults); table/markdown/json renderers; fixture integration test following the convention.
  • Depends on: 1
  • Requirements: R2
  • Test: pytest cli/tests/test_cli.py -k scenarios_command (red→green)
• [x] 3. Config contract: testing + apiSpecs
  • Extend config.schema.json; mirror in templates/config.yaml and the repo's own config.yaml; track specs/openapi/ + specs/graphql/ patterns in the manifest.
  • Depends on: none
  • Requirements: R1, R3, R4
  • Test: uv run python scripts/validate_config.py
• [x] 4. Codify conventions in the skill
  • New reference/testing.md; wire into SKILL.md (reference list, operating principles, config sections), reference/tooling.md, the design template, READMEs.
  • Depends on: 3
  • Requirements: R1, R3, R4
  • Test: npx markdownlint-cli2 "**/*.md"
• [x] 5. Decision record + spec artifacts
  • decision-014 (incl. the GraphQL-TODO resolution) + this 3-phase spec + execution log.
  • Depends on: 4
  • Requirements: R4
  • Test: npx markdownlint-cli2 "**/*.md"

Dependency graph (DAG)

1 → 2; 3 → 4 → 5 (the two chains are independent until 4 references 1–3).

Checkpoints

After tasks 2, 3 and 5: make check (ruff · pyright · schema validation · pytest · markdownlint) — the same command CI runs.