RohitKattimani/RELAY ? reverse-engineered prompt

Build me a local incident response app for DevOps teams that helps track outages, store what happened, and remember how similar incidents were solved before.

I want a React front end and a Python backend with a simple database, where I can create incidents, add symptoms, logs, and context, then analyze them with an AI assisted helper. It should search past resolved incidents, show likely root causes with source links, and suggest useful runbooks or next steps. If there is no AI key set up, it should still work with a clear retrieval only fallback.

Please include incident history, resolution notes, reopening, runbook management, and a timeline of activity and signals. It should also support importing GitHub workflow results as signals, and have a demo mode with sample incidents so I can try it right away. Make it feel polished, local first, and safe, with only advisory guidance and no ability to run commands or change infrastructure.