Ingest GitHub Issues and Pull Requests into LLM-friendly text dumps.
Live Demo: gitthread.streamlit.app
gitthread is a companion tool to gitingest that focuses on the conversational context of GitHub Issues and Pull Requests.
This project is inspired by and uses the library from gitingest. Special thanks to the gitingest team for their amazing work on codebase ingestion.
- Issue & PR Ingestion: Extract title, body, and all comments from any public GitHub issue or PR.
- Smart Linking: Automatically detects and ingests linked issues (e.g., #123 or "fixes #123") in PR descriptions.
- Repository Context: Integrates with gitingest to provide a summary or full content of the repository.
- Concurrent Processing: Optimized fetching of both thread and repository data.
- LLM-Friendly: Outputs clean Markdown formatted for easy consumption by LLMs.
- CLI & Web: Use it in your terminal or via a polished Streamlit dark-themed interface.
pip install gitthreadgitthread https://github.com/user/repo/issues/1gitthread-webFor instructions on how to deploy gitthread using Docker or Docker Compose (including Dockge), please see DOCKER.md.
MIT
