This project provides an AI-powered system for automatically analyzing codebases and generating comprehensive documentation, targeting developers and teams seeking to improve code understanding and onboarding. It leverages a multi-agent architecture and supports multiple LLM providers for enhanced code comprehension and automated documentation creation.

How It Works

The system employs a multi-agent architecture where specialized AI agents perform distinct code analysis tasks, including code structure, data flow, dependency, request flow, and API analysis. This modular approach allows for focused processing and potentially more accurate, detailed documentation. The generated documentation is then assembled into comprehensive README files, with features for concurrent processing and GitLab integration for automated workflows.

Quick Start & Requirements

• Install: Clone the repository, then use uv sync or pip install -e ..
• Prerequisites: Python 3.13, Git, and API access to an OpenAI-compatible LLM provider.
• Setup: Copy .env.sample to .env and config_example.yaml to .ai/config.yaml.
• Run: uv run src/main.py analyze --repo-path . and uv run src/main.py document --repo-path ..
• Docs: Blog Posts

Highlighted Details

• Multi-agent analysis for specialized code aspects.
• Automated README generation with configurable sections.
• GitLab integration for merge request creation.
• Concurrent processing for improved performance.
• Support for multiple OpenAI-compatible LLMs.
• Observability via OpenTelemetry and Langfuse.

Maintenance & Community

The project is built with pydantic-ai and integrates with Langfuse for observability. No specific community links or contributor details are provided in the README.

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

The project requires Python 3.13, which is a recent version and may not be universally adopted. The README does not detail performance benchmarks or specific limitations of the AI analysis.