Add RAG-based chatbot for querying ingested data

[vishwamartur]

Related to #48

Integrate a RAG-based chatbot for querying ingested GitHub repository data.

• Backend Changes:

  • Add src/gitingest/embedding.py to handle vectorization of repository content.
  • Add src/gitingest/rag_chatbot.py to manage retrieval and generation logic.
  • Add new API endpoints in src/routers/chatbot.py for chatbot communication.

• Frontend Updates:

  • Update src/templates/index.jinja and src/templates/github.jinja to include a chatbot UI with input and response display elements.

• Dependencies:

  • Update requirements.txt to include sentence-transformers, openai, langchain, and faiss-cpu.

• Testing:

  • Add test cases in src/gitingest/tests/test_embedding.py to validate embedding generation and storage.
  • Add test cases in src/gitingest/tests/test_rag_chatbot.py to validate retrieval accuracy and response quality.
  • Add test cases in src/gitingest/tests/test_chatbot_endpoints.py to validate API endpoint functionality.

[cyclotruc]

Wow that was fast, very exciting thank you

Unfortunately from my first tests sending prompts such as "only docs" or "python files please" didn't seem to change anything to the result,
Could you please provide steps to try out and see a noticeable difference?

Thanks again, I believe quick drafts are a good way to iterate