https://github.com/vinodnimbalkar/gcm
`gcm` is a command-line tool leveraging local LLMs via Ollama OR Google's Gemini API to generate Git commit messages with relevant emojis based on the Git diff.
https://github.com/vinodnimbalkar/gcm
ai aicommit commit gemini gemini-api gemma3 git olla rust
Last synced: 28 days ago
JSON representation
`gcm` is a command-line tool leveraging local LLMs via Ollama OR Google's Gemini API to generate Git commit messages with relevant emojis based on the Git diff.
- Host: GitHub
- URL: https://github.com/vinodnimbalkar/gcm
- Owner: vinodnimbalkar
- License: mit
- Created: 2025-04-19T08:48:04.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2025-05-10T03:26:08.000Z (5 months ago)
- Last Synced: 2025-05-10T04:24:11.686Z (5 months ago)
- Topics: ai, aicommit, commit, gemini, gemini-api, gemma3, git, olla, rust
- Language: Rust
- Homepage:
- Size: 61.5 KB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# GCM - AI-Powered Git Commit Message Generator
`gcm` is a command-line tool that generates meaningful Git commit messages with appropriate emojis based on your Git diff, powered by local LLMs through Ollama.
## Description
This tool pipes your Git diff into an Ollama instance running locally to generate semantically relevant commit messages that follow conventional commit formats with emoji prefixes. It's perfect for developers who want consistent, meaningful commit messages without the mental overhead of crafting them each time.
## Features
- Generates commit messages with appropriate emoji prefixes
- Analyzes Git diff content to create contextual commits
- Follows conventional commit standards
- Runs completely locally (no data sent to external APIs)
- Fast and lightweight
## Prerequisites
- [Rust](https://www.rust-lang.org/tools/install) (for building)
- [Ollama](https://ollama.ai/) running locally
- A compatible language model (default: "gemma3")
## Installation
### From Source
```sh
git clone https://github.com/vinodnimbalkar/gcm.git
cd gcm
cargo install --path .
```
### From Release
This one-line installation does not support every OS
```bash
curl -sL "https://github.com/vinodnimbalkar/gcm/releases/download/v1.0.2/gcm" | sudo tee ~/.local/bin/gcm >/dev/null && sudo chmod +x ~/.local/bin/gcm
```
### Using the Docker Image
You can directly use the pre-built Docker image from Docker Hub:
```sh
docker pull vinodnimbalkar/gcm:latest
```
Using with Gemini API via Docker
```sh
git diff | docker run -i --rm -e GEMINI_API_KEY=your_api_key_here vinodnimbalkar/gcm:latest
```
Creating a Shell Alias for Docker
```sh
# Add to .bashrc, .zshrc, etc.
alias docker-gcm='commit_message=$(git diff --staged | docker run -i --rm -e GEMINI_API_KEY=your_api_key_here vinodnimbalkar/gcm:latest); git commit -m "$(echo "$commit_message" | head -n 1)" -m "$(echo "$commit_message" | tail -n +2)"'
```
## Usage
Simply pipe your Git diff to the tool:
```sh
git diff | gcm
```
Or for staged changes:
```sh
git diff --staged | gcm
```
Then use the generated message to commit:
```sh
git commit -m "$(git diff --staged | gcm)"
```
For convenience, you might want to add an alias to your shell configuration:
```sh
# Add to .bashrc, .zshrc, etc.
alias gcm='commit_message=$(git diff --staged | gcm); git commit -m "$(echo "$commit_message" | head -n 1)" -m "$(echo "$commit_message" | tail -n +2)"'
```
## LLM Service Configuration
The tool supports two LLM services:
### 1. Ollama (Default, Local)
If no Gemini API key is set, `gcm` will use Ollama by default. Ollama must be installed and running locally.
You can configure the Ollama model using an environment variable:
```sh
export OLLAMA_MODEL=gemma3 # Default is "gemma3"
```
### 2. Gemini (Cloud API)
To use Google's Gemini API:
1. Obtain a Gemini API key from [Google AI Studio](https://makersuite.google.com/app/apikey)
2. Set the API key as an environment variable:
```sh
export GEMINI_API_KEY=your_api_key_here
```
`gcm` will automatically use Gemini when the `GEMINI_API_KEY` environment variable is set.
## Commit Message Format
```
:
```
Example:
```
feat: ✨Add user authentication
Implement OAuth2 login flow with Google and GitHub providers
```
## Supported Commit Types and Emojis
| Type | Emoji |
|-----------------------------|-------------------|
| Initial commit | 🎉 `:tada:` |
| Version tag | 🔖 `:bookmark:` |
| New feature | ✨ `:sparkles:` |
| Bugfix | 🐛 `:bug:` |
| Metadata | 📇 `:card_index:` |
| Documentation | 📚 `:books:` |
| Documenting source code | 💡 `:bulb:` |
| Performance | 🐎 `:racehorse:` |
| Cosmetic | 💄 `:lipstick:` |
| Tests | 🚨 `:rotating_light:` |
| Adding a test | ✅ `:white_check_mark:` |
| Make a test pass | ✔️ `:heavy_check_mark:` |
| General update | ⚡ `:zap:` |
| Improve format/structure | 🎨 `:art:` |
| Refactor code | 🔨 `:hammer:` |
| Removing code/files | 🔥 `:fire:` |
| Continuous Integration | 💚 `:green_heart:` |
| Security | 🔒 `:lock:` |
| Upgrading dependencies | ⬆️ `:arrow_up:` |
| Downgrading dependencies | ⬇️ `:arrow_down:` |
| Lint | 👕 `:shirt:` |
| Translation | 👽 `:alien:` |
| Text | ✏️ `:pencil:` |
| Critical hotfix | 🚑 `:ambulance:` |
| Deploying stuff | 🚀 `:rocket:` |
| Fixing on MacOS | 🍎 `:apple:` |
| Fixing on Linux | 🐧 `:penguin:` |
| Fixing on Windows | 🏁 `:checkered_flag:` |
| Work in progress | 🚧 `:construction:` |
| Adding CI build system | 👷 `:construction_worker:` |
| Analytics or tracking code | 📈 `:chart_with_upwards_trend:` |
| Removing a dependency | ➖ `:heavy_minus_sign:` |
| Adding a dependency | ➕ `:heavy_plus_sign:` |
| Docker | 🐳 `:whale:` |
| Configuration files | 🔧 `:wrench:` |
| Package.json in JS | 📦 `:package:` |
| Merging branches | 🔀 `:twisted_rightwards_arrows:` |
| Bad code / needs improvement| 💩 `:hankey:` |
| Reverting changes | ⏪ `:rewind:` |
| Breaking changes | 💥 `:boom:` |
| Code review changes | 👌 `:ok_hand:` |
| Accessibility | ♿ `:wheelchair:` |
| Move/rename repository | 🚚 `:truck:` |
## Configuration
The tool currently uses "gemma3" as the default model. You can modify the source code to use different Ollama models.
## License
MIT License
## Contributing
Contributions are welcome! Please feel free to submit a Pull Request. Contributions are very welcome!! 😄