Acceptance Test

The result is a passing (or honestly failing) Playwright suite in which every test maps to an acceptance criterion the user has approved. Criteria come from the user, not from the code. The code only shows what exists, not what should exist.

Workflow

Copy this checklist and tick each phase off:

- [ ] 1. Learn: behavior inventory built from code + docs
- [ ] 2. Grill: criteria agreed, user confirmed shared understanding
- [ ] 3. Record: tests/acceptance/CRITERIA.md written
- [ ] 4. Set up: Playwright installed, app starts via webServer
- [ ] 5. Write: one test per criterion, IDs in titles
- [ ] 6. Run + triage: every failure classified
- [ ] 7. Report: criterion → result table