gcommit 🚀

gcommit is a command-line tool that leverages Google Cloud's AI models (specifically Gemini) to generate concise, well-formatted Git commit messages that adhere to Conventional Commits standards. It automates the creation of commit messages based on your staged changes.

How to Use

1. Stage your changes: In your Git repository, add the files you want to include in the commit:

   git add .
   # or
   git add <specific_files>

2. Run gcommit: Make sure you have installed and configured gcommit.

   gcommit

3. Confirm the message: gcommit will display the generated commit message and ask for your confirmation.

   ✨ Generated commit message:
   
   feat: PROJ-123 implement user authentication flow
   
   Use this commit message? [Y/n]:
   

   Press Y (or Enter) to proceed with the commit, or n to cancel.

4. Forward arguments to git commit: You can pass arguments directly to git commit after gcommit. For example, to sign a commit with GPG:

   gcommit -S

   Or to add a longer description (opening the editor):

   gcommit -v

Key Features

• ✨ AI-Generated Commit Messages: Uses Google Cloud's Gemini models to create intelligent commit messages.
• 🏷️ Ticket ID Extraction: Automatically extracts ticket IDs (e.g., PROJ-123) from the branch name and prepends them to the message.
• 🧠 Context-Aware: Considers recent commit history and the files being modified.
• convention Conventional Commits: Generates messages that follow the Conventional Commits standard.
• 📏 Concise and Formatted: Produces single-line messages, under 72 characters, in the imperative mood.
• 🔧 Configurable: Set your Google Cloud project, region, and model ID via environment variables.
• 💬 Interactive Confirmation: Asks for your approval before committing.
• passthrough Argument Forwarding: Allows you to forward additional arguments to the git commit command (e.g., gcommit -S to sign the commit).
• 🐛 Debug Mode: Includes a --debug mode to inspect the prompt sent to the AI and its response.

Prerequisites

Before installing gcommit, ensure you have:

• Git: Installed and configured.
• jq: Installed (e.g., sudo apt install jq or brew install jq).
• Google Cloud SDK (gcloud): Installed and configured. You must be authenticated (gcloud auth login) and have a default project configured (gcloud config set project YOUR_PROJECT_ID).
• Google Cloud Project:
  • The Vertex AI API must be enabled.
  • Billing must be configured for the project.

Installation

1. Clone the repository (if you haven't already):

   git clone <YOUR_REPOSITORY_URL>
   cd <REPOSITORY_FOLDER_NAME> # e.g., cd gcommit

2. Run the installation script:

   bash install.sh

   This script will copy gcommit to ~/.local/bin and make it executable. Follow the instructions provided by the script if ~/.local/bin is not in your PATH.

Configuration

gcommit uses the following environment variables:

• GGIT_PROJECT: (Required) Your Google Cloud project ID.
• GGIT_LOCATION: (Optional) The region for AI API calls. Default: europe-west8.
• GGIT_MODEL_ID: (Optional) The Gemini model ID to use. Default: gemini-2.0-flash-001.

Add these variables to your shell's configuration file (e.g., ~/.bashrc, ~/.zshrc):

export GGIT_PROJECT="YOUR_GCP_PROJECT_ID"
# export GGIT_LOCATION="your-region" # Optional
# export GGIT_MODEL_ID="your-gemini-model" # Optional