pankaj-2708/Research_Paper_Vectorless_RAG — reverse-engineered prompt

Build me a simple Python tool for research papers that uses a vectorless RAG approach instead of embeddings. I want to be able to give it one or more paper files or pasted text, then ask normal language questions and get useful answers grounded in the actual paper content. It should be good for things like summarizing a paper, finding where a method or result is discussed, comparing sections, and pulling out relevant passages before answering.

Keep it practical and easy to run locally, with a clean command line flow or a very small local app if that makes more sense. Please structure it so the retrieval logic is clear and the answers stay tied to the source text, not generic model guesses. Add basic setup instructions, an example of how to use it, and a sample question flow so I can test it right away.

If anything is unclear, use the repo name as the intent and look up current docs online if you need to.