anjaney24102-star/S-P- ? reverse-engineered prompt

Build me a local financial risk monitoring app that can take in news, social posts, or manual entries, analyze them, and turn them into clear risk signals.

I want an API that accepts events, stores them, de duplicates repeated stories, and lets me fetch recent events, analyses, signals, entities, and warnings. The analysis should work fully offline with simple rule based logic, giving sentiment, event type, impact score, key entities, and evidence, then generate explainable risk signals from that. It should also resolve company and entity names so things like Apple, Apple Inc., and AAPL are treated as the same entity when the match is strong enough.

Please include early warning detection that compares recent activity to a baseline and returns a simple severity like WATCH, ELEVATED, HIGH, or CRITICAL. I’d also like a small working frontend or demo pages if that fits the project, plus tests and a clean health endpoint. Keep the data format stable and easy to use, and make it run locally with a simple setup.