Gitingest is a Git repository analysis and conversion service that transforms code repositories into structured plain-text summaries optimized for large language model consumption. It provides HTTP API endpoints and Python functions to integrate repository processing into AI pipelines and applications, with S3-compatible storage for persisting and retrieving generated digests. The service is packaged as a Docker container with all dependencies bundled for consistent deployment across environments.
The project distinguishes itself through asynchronous processing of multiple repositories concurrently using a worker pool for high throughput, and the ability to stream repository digests via HTTP endpoints for real-time consumption in AI workflows. It includes comprehensive monitoring through Prometheus metrics and Sentry exception tracking to track performance and detect issues in digest generation workflows, while maintaining security best practices through non-root execution and vulnerability reporting mechanisms.
Gitingest offers a command-line interface for triggering repository conversions without programming, along with development tools for faster iteration including hot reload support. The system is designed to preserve code hierarchy and snippets during repository conversion, ensuring the generated digests maintain the structural integrity of the original codebase for more accurate AI analysis.
The project is implemented in Python and can be deployed as a containerized service in CI/CD pipelines with all dependencies included.
