dtquocbao/ai-market — reverse-engineered prompt

Build me a local AI market analysis app that feels like a real trading research dashboard, not a toy demo. I want a web UI and backend that pull live market data, show a clean overview of market conditions, and give AI generated insights I can actually read and compare. It should include multiple specialized agents that analyze things like news and events, strategy ideas, technical signals, forecasts for day and swing trading, risk analysis, ticker discovery, and overall agent performance.

I also want symbol management with saved data in a real database so scans, forecasts, and reports don’t disappear after refresh. The dashboard should have sector based market intelligence for tech, finance, healthcare, and retail, plus confidence scores, explanations, analytics, alerts, status pages, and simple settings. Make the API endpoints and frontend actually work together and make sure the app runs locally with the existing Docker and database setup.

Please wire up anything missing, fix broken routes or persistence issues, and make it feel polished and usable. Look up current docs online if you need to.