bts-testlisted

# Test: Generate + Execute + Fix Loop Generate and run tests for recipe: $ARGUMENTS ## Settings Read `.bts/config/settings.yaml` for project-specific limits. ## Prerequisites 0. **Resolve recipe ID**: If `$ARGUMENTS` is empty, run `bts recipe status` to find the active recipe. Use its ID for all `{id}` references below. 1. Verify tasks.json exists and implementation is done: ```bash ls .bts/specs/recipes/{id}/tasks.json ``` If not found → "Run /implement first." 2. Check tasks.json: all tasks should be `done` or `skipped` (no `pending` or `in_progress`). If incomplete → "Implementation not finished. Complete /implement first." ## Resume If test-results.json already exists: - Skip Steps 1-2 (test scenarios already extracted, test code already generated) - Go directly to Step 3 (always re-run tests — code may have changed since last run) - If status was `"fail"` → read previous failures for context before re-running If test-results.json does not exist → start from Step 1 ## Step 1: Extract Test Scenarios 1. Read `.bts/specs/recipes/{id}/final.md` 2. Find all test scenarios defined in the spec: - Happy path scenarios - Error path scenarios - Edge cases - Integration scenarios 3. **Read simulations/*.md (Phase 13)**: collect every scenario id referenced in simulation files. These are the ids that tests MUST link via `bts:scenario` tags. `bts check test-coverage` enforces this after Step 2.5 runs. 4. Read existing test files in th