GitSearch

A command-line tool that uses OpenAI's embeddings and GPT models to help you find relevant commits in your git repository based on natural language queries.

Features

• Semantic search through git commit history using OpenAI embeddings
• Automatic detection of git repository
• Configurable number of relevant commits to analyze
• Natural language querying of commit history

Installation

1. Clone the repository:

git clone https://github.com/trulsmoholt/gitsearch.git
cd gitsearch

2. Install the required packages:

pip install -r requirements.txt

3. Install the package locally:

pip install --user .

Configuration

Set your OpenAI API key as an environment variable:

export OPENAI_API_KEY="your-api-key-here"

Or add it to your ~/.zshrc or ~/.bashrc:

echo 'export OPENAI_API_KEY="your-api-key-here"' >> ~/.zshrc
source ~/.zshrc

Usage

Basic Usage

From within a git repository:

gitsearch "fix login button not working"

Command Line Options

• --repo: Path to git repository (defaults to current directory)
• --api-key: OpenAI API key (defaults to OPENAI_API_KEY environment variable)
• --model: OpenAI model to use (default: "gpt-4")
• --days: Number of days to look back (default: 30)
• --top-k: Number of most relevant commits to analyze (default: 5)

Examples:

# Search in specific repository
gitsearch "fix login button" --repo /path/to/repo

# Look back 60 days
gitsearch "fix login button" --days 60

# Analyze top 3 most relevant commits
gitsearch "fix login button" --top-k 3

# Use specific API key
gitsearch "fix login button" --api-key your-api-key-here

Development

For development and testing, you can run the script directly:

python gitsearch.py "your query"

Requirements

• Python 3.7+
• OpenAI API key
• Git repository

Dependencies

• openai
• gitpython
• numpy