This interactive web tool helps developers explore and understand GitHub repositories more effectively. By combining a file tree visualisation with LLM-powered analysis, it enables users to grasp project structure, purpose, and individual file functionality without manually reading through the codebase.
- Paste a GitHub repo URL to visualise its file structure.
- Get project architecture and overview of the project.
- Analyse any file in the repository to understand its purpose, components, and functionality.
- It also provides insights into areas where you can contribute to the project.
- FastAPI backend with AI integration (8x more powerful than Mistral 7B).
| File | Description |
|---|---|
main.py |
FastAPI app with CORS and three API routes for analysis |
repo_handler.py |
Handles GitHub API interactions and builds repo file tree |
llm_integration.py |
Sends repo/code data to Together.ai and parses the response |
index.html |
Frontend UI with tree view and analysis controls |
requirements.txt |
Python dependencies for the backend |
- Python 3.10+
- GitHub Personal Access Token (
GITHUB_PAT) - Together.ai API key (
TOGETHER_API_KEY) - FastAPI (Backend framework)
- Uvicorn (Server for FastAPI)
- Requests (Module for handling requests)
- Python-dotenv (for .env file)
- Frontend: Vercel (or any static host)
- Backend: Render (or any FastAPI-compatible host)
-
- Clone the repo and install dependencies
pip install -r requirements.txt
-
- Create .env file
TOGETHER_API_KEY=your_together_api_key GITHUB_PAT=your_github_pat
-
- Run the backend
uvicorn main:app --reload
-
- Run frontend on browser and check API calls in Swagger
Access Swagger at /docs route at port 8000 or Localhost