RepoMind

RepoMind is an AI-powered tool that lets users query and understand GitHub repositories using natural language. It leverages Retrieval-Augmented Generation (RAG), OpenAI models, and vector embeddings to deliver context-aware responses grounded in source code.

🌐 Live Demo

Link

🚀 Features

• Fetches GitHub commits and stores relevant data in PostgreSQL
• Uses LangChain, OpenAI, and Gemini to generate summaries and answer queries
• Embeds files with vector representations for accurate retrieval
• Implements RAG workflows for deep understanding of repository context
• Real-time, fully typesafe communication between frontend and backend using tRPC
• Built with Vite for a lightning-fast frontend dev experience

🛠 Tech Stack

• Frontend: Vite, Next.js, TypeScript, Tailwind CSS
• Backend: tRPC, Prisma, PostgreSQL
• AI/ML: OpenAI API, Gemini API, LangChain, custom RAG pipeline
• Deployment: Vercel

Prerequisites

• Node.js (v14+)
• Python 3.8+
• PostgreSQL
• OpenAI API Key

🤝 Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Environment variables

• DATABASE_URL="postgresql://<username>:<password>@<host>/<database>?sslmode=require"
• GITHUB_TOKEN=<your_github_token>
• GEMINI_API_KEY=<your_gemini_api_key>
• OPENAI_API_KEY=<your_openai_api_key>