Git Analyzer

A simple command-line interface (CLI) tool that leverages AI agents to analyze Git commits for potential security and performance risks.

Features

• Commit Summarization: Provides a concise summary of the changes in a given commit.
• Risk Analysis: Identifies potential security vulnerabilities and performance bottlenecks introduced in a commit.
• Code Review: Summaries and identifies potential vulnerabilities in repo diff before committing.
• Extensible LLM Support: Easily configure different Language Model (LLM) providers through CrewAI.
• File Content Retrieval: Agents can fetch the full content of a file for more in-depth analysis.
• External Information Search: Agents can search for information about identified risks and provide relevant links.

Installation

1. Clone the repository:

   git clone https://github.com/chains229/git-analyzer.git
   cd git-analyzer

2. Install dependencies using uv:

   uv pip install -r requirements.txt

3. Set up your environment variables:

   • Create a .env file from the .env.example:

     cp .env.example .env

   • Open the .env file and add your Azure OpenAI API key.
   • You can custom the default model in source/main.py and the API key to use your preferred model.

Usage

First, enter the virtual environment of this repo

-m venv .venv
source .venv/bin/activate  # On Linux/macOS
.\.venv\Scripts\activate   # On Windows

Then, cd to the repo that contains the commit you want to analyze (i know it sucks). The main command to run the analyzer is ga.

• To summary a commit:

ga summary --commit <your-commit-id>

• To analyze performance/security risk of a commit:

ga summary --commit <your-commit-id>

• To review uncommitted code:

ga review

Optional argument

• --model: Choose the LLM you want.
• --output-dir: You can save the result into a .txt file by setting this argument. (dont think it's working)