GitHub Sentinel is an AI agent designed for intelligent information retrieval and high-value content mining in the era of large language models (LLMs). It targets users who need to frequently access and analyze large volumes of information, such as open-source enthusiasts, individual developers, and investors. The agent automates the tracking and analysis of GitHub repository updates and can be extended to other information sources like Hacker News, providing comprehensive data mining and analysis capabilities.

How It Works

GitHub Sentinel leverages a subscription-based model to track specified GitHub repositories and can also ingest data from Hacker News. It automatically retrieves updates (commits, issues, pull requests) and uses LLMs (OpenAI or Ollama) to generate natural language reports on project progress or trending topics. The system supports scheduled tasks for continuous monitoring and offers a Gradio-based GUI for user-friendly interaction, alongside Docker support for streamlined deployment.

Quick Start & Requirements

• Install dependencies: pip install -r requirements.txt
• Configuration: Edit config.json for GitHub token, email settings, subscriptions, LLM configuration (OpenAI/Ollama), and report types. Environment variables can be used for sensitive information like GitHub tokens and email passwords.
• Running modes:
  • Command-line tool: python src/command_tool.py
  • Background service: ./daemon_control.sh start
  • Gradio server: python src/gradio_server.py (default access: http://localhost:7860)
• Ollama: Requires separate installation and running of Ollama service (ollama serve) for local LLM integration.
• Documentation: GitHub Sentinel

Highlighted Details

• Supports both OpenAI and Ollama for LLM-powered report generation.
• Offers multiple operational modes: CLI, background daemon, and Gradio web UI.
• Includes automated unit tests and a validate_tests.sh script for CI/CD integration.
• Docker support with Dockerfile and build_image.sh for reproducible builds.

Maintenance & Community

• Project maintained by Django Peng. Contact: pjt73651@email.com.

Licensing & Compatibility

• Licensed under the Apache-2.0 license. Permissive for commercial use and closed-source linking.

Limitations & Caveats

The project relies on external API keys (GitHub, email) and LLM services, which may incur costs or require specific setup. Configuration requires careful editing of config.json and potentially setting environment variables.