samyak1997/biaa-healthcare ? reverse-engineered prompt

Reverse engineered prompt

Build me a small healthcare demo app that rewrites a hospital discharge summary into plain language a patient can understand, but never invents advice or changes the meaning.

The app should take a clinician written summary, strip out any patient identifiers before anything goes to an AI model, then generate a patient friendly version, review it for safety, and if the review finds problems it should rewrite it again a few times before giving up and flagging it for human review. I want a clear trace of each step so someone can see what happened in the background.

Keep it simple, no database, no vector store, no heavy ML stuff. Use one container, a Groq API key, and make sure the logic fails closed if redaction or review does not work. Use synthetic data only and make the interface easy for non technical users. If you need current library details, look up the latest docs online.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab