majiayu000 Test From Docs
Use documentation as the authoritative test blueprint. Tests verify intended behavior (from docs), not just current behavior.
When to Use
- After implementing a feature or fix
- Before marking work as complete
- When user asks "does this work?"
- As part of pre-PR validation
Risk-Based Test Selection
Calculate Risk Score
| Factor | Points |
|---|---|
| Touches auth/RLS | +3 |
| Touches billing | +3 |
| Touches schema | +2 |
| Touches AI/prompts | +2 |
| Multi-feature change | +2 |
| User-facing workflow | +1 |
| New code (not fix) | +1 |
Select Test Depth
| Score | Depth | What to Run |
|---|---|---|
| 0-2 | Light | Static + primary feature checklist |
| 3-5 | Standard | Light + impacted features + 1 workflow |
| 6-8 | Thorough | Standard + browser verification + regression |
| 9+ | Comprehensive | Full workflow suite + edge cases |
Test Phases
- Static Analysis —
npx tsc --noEmit && npm run lint - Feature Checklist — Run items from feature doc's Testing Checklist
- Integration — Verify cross-feature integration points
- Browser — Visual verification with Playwright MCP
- Workflow Smoke — One complete user journey
- Regression — Verify coupled features still work
Output
## Test Report — [feature] — [date]
**Risk Score**: [n] ([depth level])
| Phase | Passed | Failed |
|-------|--------|--------|
| Static | [n] | [n] |
| Feature | [n] | [n] |
| Browser | [n] | [n] |
| Workflow | [n] | [n] |
**Overall**: ✅ PASS / ❌ FAIL
### Failures
[Test name, expected, actual, evidence]
**Ready for Merge**: [Yes/No]
Related
- Phase details: See reference/test-phases.md
- Templates by change type: See reference/test-templates.md
- Browser testing commands: See reference/browser-testing.md
- For comprehensive testing:
/spawn-test-agent