Generate intelligent git commit messages using your local Ollama instance. No API keys, completely free, and runs offline!

• 🤖 AI-Powered: Uses local Ollama models to generate commit messages
• 🔒 Privacy First: Everything runs locally, no data sent to external APIs
• 🎯 Multiple Styles: Conventional commits, semantic, or detailed formats
• 💰 Free: No API costs, uses your local Ollama instance
• ⚡ Fast: Quick generation with local models
• 🎨 Interactive: Review, regenerate, or edit messages before committing
• 🌐 Offline: Works completely offline

pip install ollama-git-commit

git clone https://github.com/himanshu231204/ai-commit.git
cd ai-commit
pip install -e .

Before using Ollama Git Commit, you need:

1. Python 3.8+

   python --version

2. Git

   git --version

3. Ollama - Install Ollama

   # Install Ollama (macOS/Linux)
   curl -fsSL https://ollama.ai/install.sh | sh
   
   # Pull a model (e.g., llama2)
   ollama pull llama2
   
   # Start Ollama server
   ollama serve

1. Stage your changes:

   git add .

2. Generate commit message:

   ai-commit

3. Review and choose:

   • y - Accept and commit
   • r - Regenerate message
   • e - Edit message manually
   • n - Cancel

# Make some changes to your code
echo "print('Hello World')" > hello.py

# Stage the changes
git add hello.py

# Generate AI commit message
ai-commit

# Output:
╔═══════════════════════════════════════════╗
║         🤖 AI Commit Message Tool         ║
║      Powered by Local Ollama 🦙           ║
╚═══════════════════════════════════════════╝

🔍 Checking Ollama server...
✓ Ollama server is running

📊 Changes:
  + 1 lines added
  - 0 lines removed

🤖 Generating commit message...

Generated Commit Message:
──────────────────────────────────────────────────
feat: add hello world script
──────────────────────────────────────────────────

Options:
  y - Accept and commit
  r - Regenerate message
  e - Edit message
  n - Cancel

If your Ollama server is running on a different host/port, edit ai_commit.py:

ollama = OllamaClient(base_url="http://192.168.1.100:11434")

ollama = OllamaClient(model="codellama")  # or "mistral", "llama2", etc.

The tool supports three commit message styles:

1. Conventional Commits (default):

   feat(auth): add user login functionality
   

2. Semantic:

   Add user login functionality
   

3. Detailed:

   Add user authentication system
   
   - Implement JWT-based authentication
   - Add login and logout endpoints
   - Create user session management
   

AI Commit automatically selects the best available model on your system, prioritizing lighter and faster models for optimal performance.

1. phi - Smallest, fastest (Recommended for quick commits)
2. mistral - Great balance of speed and quality
3. qwen - Fast alternative
4. gemma - Google's lightweight model
5. codellama - Optimized for code
6. llama2 - Reliable but slower
7. llama3 - Most powerful but slowest

When you run ai-commit, the tool will:

1. Check all installed Ollama models
2. Automatically select the lightest/fastest available
3. Display which model was chosen
4. Generate your commit message

$ ai-commit

╔═══════════════════════════════════════════╗
║         🤖 AI Commit Message Tool         ║
║      Powered by Local Ollama 🦙           ║
╚═══════════════════════════════════════════╝

🔍 Checking Ollama server...
✓ Ollama server is running

Available models (3):
  • phi:latest
  • mistral:latest
  • llama2:latest

✓ Auto-selected: phi:latest
  (Prioritizing lighter/faster models)

📊 Changes:
  + 8 lines added
  - 0 lines removed

🤖 Generating commit message...

For best performance, install a lightweight model:

# Fastest (Recommended)
ollama pull phi

# Good balance
ollama pull mistral

# For code-heavy projects
ollama pull codellama

If you want to use a specific model, you can set an environment variable:

Windows:

set OLLAMA_MODEL=mistral
ai-commit

Linux/Mac:

export OLLAMA_MODEL=mistral
ai-commit

$ git add new_feature.py
$ ai-commit

Generated: feat: add user profile management feature

$ git add bug_fix.py
$ ai-commit

Generated: fix: resolve null pointer exception in login

$ git add README.md
$ ai-commit

Generated: docs: update installation instructions

Error: Cannot connect to Ollama
Solution: Start Ollama server with `ollama serve`

Error: No staged changes found
Solution: Stage your changes with `git add <files>`

Error: Model not found
Solution: Pull the model with `ollama pull llama2`

Contributions are what make the open-source community amazing! Any contributions you make are greatly appreciated.

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes
4. Run tests (when available)
5. Commit using ai-commit! 😄
6. Push to your branch (git push origin feature/amazing-feature)
7. Open a Pull Request

See CONTRIBUTING.md for detailed guidelines.

This project is licensed under the MIT License - see the LICENSE file for details.

• Ollama - For making local LLMs accessible
• Git - The best version control system
• Python - For being awesome
• You - For using this tool!

• Initial release on PyPI
• Configuration file support (.ai-commit.yml)
• More commit message formats
• Interactive model selection
• Emoji support in commits 🎉
• Multiple language support
• Git hooks integration
• VSCode extension
• Custom prompt templates
• Commit message history
• Auto-detect commit type from files
• Batch commit support
• Integration with GitHub CLI

Himanshu Kumar

• 🌐 GitHub: @himanshu231204
• 💼 LinkedIn: himanshu231204
• 🐦 Twitter/X: @himanshu231204
• 📧 Email: [email protected]

If you find this project helpful, please consider:

• ⭐ Starring the repository
• 🐛 Reporting bugs
• 💡 Suggesting new features
• 🔀 Contributing code
• ☕ Buy me a coffee
• 💝 Sponsor on GitHub

• PyPI Package: https://pypi.org/project/ollama-git-commit/
• Documentation: GitHub Wiki
• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Releases: GitHub Releases
• Changelog: CHANGELOG.md

Give a ⭐️ if this project helped you!

Made with ❤️ by Himanshu Kumar