git-pulse

README

GitPulse

Go beyond Lines of Code. GitPulse uses LLMs to understand the what and why behind every commit, offering meaningful daily reports for engineering teams.

Traditional developer productivity metrics like Lines of Code (LOC) or commit frequency are often misleading. They fail to capture the context, complexity, and impact of a developer's work. Was it a critical bug fix, a complex refactor, or just dependency updates?

GitPulse addresses this challenge by leveraging the power of Large Language Models (LLMs). It connects to your GitHub or GitLab repositories, analyzes the code changes within a 24-hour window, and generates a comprehensive report for each contributor. This report includes a natural language summary of their work and a multi-faceted metric, the Daily Contribution Index (DCI), which evaluates code changes based on impact and complexity, not just volume.

Key Features

• 🤖 LLM-Powered Analysis: Generates qualitative summaries of daily work (e.g., "Implemented user authentication module," "Refactored the caching logic").
• 📈 Beyond-LOC Metrics: Calculates a configurable Daily Contribution Index (DCI) based on feature impact, code complexity, and tech debt reduction.
• 🔌 Broad Git Support: Integrates seamlessly with GitHub and GitLab, supporting both cloud and self-hosted instances.
• 🚀 Automated Daily Reports: Delivers scheduled, easy-to-read reports directly to DingTalk (with an extensible architecture for Slack, etc.).
• ⚙️ Simple Web UI: An intuitive interface for managing teams, repositories, and user mappings.
• 🌐 Open Source: Built with Go and MongoDB. Licensed under MIT.

Getting Started

(Coming soon...)

Contributing

(Coming soon...)