Git Commit AI

A VS Code extension that suggests meaningful Git commit messages based on your staged changes using AI-powered generation via the Hugging Face Inference API.

Features

• Git Integration: Analyzes staged changes with git diff --staged.
• AI-Powered Suggestions: Generates commit messages using a free Hugging Face model (e.g., Mixtral-8x7B-Instruct).
• User Interface: Displays suggestions in a VS Code Quick Pick modal, allowing you to approve, edit, or regenerate messages.
• Commit Execution: Commits directly with git commit -m when a message is approved.
• Customization: Supports different commit styles (Conventional Commits, casual, detailed) and API token configuration.
• Error Handling: Provides feedback for Git or API issues.

Installation

1. Install from VS Code Marketplace (once published):

   • Search for Git Commit AI in the Extensions view (Ctrl+Shift+X or Cmd+Shift+X).
   • Click Install.

2. Manual Installation (for development):

   • Clone this repository:

     git clone https://github.com/yourusername/git-commit-ai.git
     cd git-commit-ai
     

   • Install dependencies:

     npm install
     

   • Open in VS Code and press F5 to run in Extension Development Host mode.

Prerequisites

• Git: Installed and available in your PATH (run git --version to check).
• Hugging Face API Token: A free token from huggingface.co with Read access.

Usage

1. Stage Changes:

   • Make changes in your Git repository and stage them:

     git add .
     

2. Generate Commit Message:

   • Open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P).
   • Run Suggest Commit Message.
   • A Quick Pick menu will appear with:
     • A suggested commit message.
     • Options to "Regenerate" or "Custom" (write your own).

3. Approve or Edit:

   • Select the suggestion and press Enter to commit.
   • Choose "Custom" to edit the message in an input box.
   • Select "Regenerate" to get a new suggestion.

4. Commit:

   • The extension executes git commit -m "<message>" and shows a success message.

Configuration

Customize the extension via VS Code settings:

• gitCommitAI.apiToken:

  • Your Hugging Face API token.
  • Default: "" (empty, must be set).
  • How to set: Open Settings (Ctrl+,), search for gitCommitAI.apiToken, and paste your token.

• gitCommitAI.commitStyle:

  • Style of generated commit messages.
  • Options: conventional, casual, detailed.
  • Default: conventional.
  • Examples:
    • conventional: feat: add new feature
    • casual: Yo, added some cool stuff!
    • detailed: Updated the button text from "CV" to "Download CV" for clarity

Example

1. Stage a change:

   echo "Download CV" > page.tsx
   git add page.tsx
   

2. Run Suggest Commit Message.
3. See a suggestion like:
   • Casual: Changed the CV button text, yo!
4. Press Enter to commit.

Troubleshooting

• "No staged changes found":

  • Ensure you’ve staged changes with git add.
  • Verify you’re in a Git repository folder (run git status).

• "AI API error: 401 Unauthorized":

  • Check your Hugging Face API token in settings.

• "AI API error: 429 Too Many Requests":

  • You’ve hit the free tier rate limit. Wait or upgrade to a Hugging Face PRO account.

• No suggestions:

  • Confirm internet connectivity and token validity.
  • Check the Debug Console (F5 mode) for errors.

Development

Prerequisites

• Node.js and npm
• VS Code
• yo and generator-code:

  npm install -g yo generator-code
  

Setup

1. Scaffold the extension:

   yo code
   

2. Install dependencies:

   npm install --save axios @types/node child-process-promise
   

3. Run in debug mode:
   • Open in VS Code, press F5.

Build and Publish

1. Package:

   npm install -g vsce
   vsce package
   

2. Publish:
   • Get a Personal Access Token from Visual Studio Marketplace.
   • Run:

     vsce publish --pat <your-token>
     

Limitations

• Requires an internet connection for API calls.
• Free Hugging Face API has rate limits (~30 requests/hour).
• Single-root workspaces only (multi-root support not implemented).

Contributing

Feel free to submit issues or pull requests on GitHub.

Acknowledgments

• Built with VS Code API.
• Powered by Hugging Face Inference API.