Codebase Genius 🧠

An AI-powered multi-agent system that automatically generates high-quality documentation for any software repository.

Features

• 🔍 Automatic Repository Analysis: Clone and analyze any GitHub repository
• 🗺️ Smart Code Mapping: Generate file trees and dependency graphs
• 🧬 Code Context Graphs: Build relationships between functions, classes, and modules
• 📄 Documentation Generation: Create comprehensive markdown documentation
• 🎯 Multi-Agent System: Specialized agents for mapping, analyzing, and documenting

Architecture

• Code Genius (Supervisor): Orchestrates the entire workflow
• Repo Mapper: Explores repository structure and creates file trees
• Code Analyzer: Parses code and builds Code Context Graphs (CCG)
• DocGenie: Synthesizes documentation from analyzed data

Setup

1. Clone this repository
2. Create virtual environment: python3 -m venv venv
3. Activate: source venv/bin/activate
4. Install dependencies: pip install -r requirements.txt
5. Set up .env with your API keys
6. Run: jac serve backend/main.jac

Usage

# Start the backend server
jac serve backend/main.jac

# In another terminal, use the API
curl -X POST http://localhost:8000/walker/analyze_repo \
  -H "Content-Type: application/json" \
  -d '{"repo_url": "https://github.com/user/repo"}'

Technology Stack

• Jac: Multi-agent orchestration
• Tree-sitter: Code parsing
• FastAPI: REST API
• OpenAI/Gemini: LLM reasoning
• GitPython: Repository operations

License

MIT