Sneh14/ai-native-qa-automation — reverse-engineered prompt

Build me a small but real ticket booking app called EventHub, with a web frontend, an API, user login, events, bookings, booking details, and a simple admin area for managing events. Then add an AI powered QA workflow on top of it that feels like four specialist helpers working in sequence. One should generate test scenarios from the feature and business rules, one should decide the right level for each test, one should write and run the tests in a real browser and keep fixing them until they pass, and one should review the finished tests for quality.

I want the tests spread sensibly across unit, API, component, and end to end coverage, not just heavy UI testing. Make it use the app’s real source code, selectors, API behavior, and existing tests as context so the output is grounded. Cover important rules like booking reference format, seat count changes, refund only for quantity one, cross user access being blocked, and booking limits with oldest items pruned first.

If anything is unclear, look up current docs online and make reasonable choices.