Chirag514/RAG-Based-LLM — reverse-engineered prompt

Reverse engineered prompt

Build me a simple Python app for question answering over my own content. I want a chat style interface where I can upload PDF files or paste in a website URL, have the app read and index that content, and then ask questions about it in plain English.

The answers should be based on the uploaded documents or site content, not just general model knowledge. Please use a retrieval based approach so it finds the most relevant chunks first, then generates an answer. I also want the response to show which source text or page it came from so I can trust it.

Keep it fast and easy to run locally, with a clean interactive UI. It should support using a Groq API key from an environment file and start with one command. If anything is unclear, check the current docs online and wire up the needed pieces so the whole flow works end to end.

Want more depth? Deep Reverse