https://github.com/osama-yusuf/gitconvit
GitConvit: AI Conventional Commit Msg Generator.
https://github.com/osama-yusuf/gitconvit
bash-script conventional-commits git gum ollama python-script
Last synced: 4 months ago
JSON representation
GitConvit: AI Conventional Commit Msg Generator.
- Host: GitHub
- URL: https://github.com/osama-yusuf/gitconvit
- Owner: Osama-Yusuf
- License: mit
- Created: 2024-08-02T12:51:55.000Z (10 months ago)
- Default Branch: main
- Last Pushed: 2024-08-16T11:43:48.000Z (10 months ago)
- Last Synced: 2025-01-31T06:51:18.867Z (4 months ago)
- Topics: bash-script, conventional-commits, git, gum, ollama, python-script
- Language: Shell
- Homepage:
- Size: 2.19 MB
- Stars: 3
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
GitConvit, an AI-powered tool that generates conventional commit messages with emojis based on the changes in your Git repository.
---
This tool leverages the `llama3:latest` model from `ollama` to create commit messages that follow conventional commit standards, ensuring that your commit history is informative and well-structured.
## Features
- Automatically detects changes in your Git repository.
- Generates conventional commit messages with appropriate emojis.
- Interactive prompt to confirm or regenerate the commit message along other options.
- Ensures that all changes are committed with a single, informative message.
## Emojis and Categories
The following categories and emojis are used for generating commit messages:
- `docs`: 📝
- `feat`: ✨
- `fix`: 🐛
- `style`: 🎨
- `refactor`: 🔨
- `chore`: 🚀
- `config`: ⚙️
## Screenshots
## Prerequisites
1. **Install `ollama`**:
- Download and install `ollama` from [ollama's official website](https://www.ollama.com/download).
- Ensure that `ollama` is running and accessible at `http://localhost:11434`.
- Pull the `latest` version of the `llama3` model. with the following command:
```bash
ollama pull llama3:latest
```
2. **Install `gum` (Optional for Enhanced UI/UX)**:
- If you want to use the `gum` enhanced script:
```bash
# macOS or Linux
brew install gum
# Arch Linux (btw)
pacman -S gum
# Windows (via WinGet or Scoop)
winget install charmbracelet.gum
scoop install charm-gum
```
- More installation methods [here](https://github.com/charmbracelet/gum#Installation).
## Why Three Scripts?
To cater to different user preferences and environments, we provide three versions of the script:
1. **Python Script (`script.py`)**:
- For users who prefer Python and want a straightforward approach without additional dependencies.
2. **Bash Script without Dependencies (`script.sh`)**:
- For users who prefer a bash script without any additional dependencies. This version has a basic user interface and functionality.
3. **Bash Script with `gum` (`gumscript.sh`)**:
- For users who want a significantly enhanced user experience with a modern, interactive UI. This version requires the `gum` tool to be installed.
You can choose the version that best fits your needs and preferences.
## Installation
1. **Clone the Repository**:
```bash
git clone https://github.com/Osama-Yusuf/GitConvit.git
cd GitConvit
```
2. **Make the Script Executable**:
```bash
chmod +x gumscript.sh
```
3. **Install Dependencies**:
- Ensure you have `jq` installed for processing JSON responses:
```bash
sudo apt-get install jq
```
4. **Add the Script to your bin directory**:
```bash
sudo cp gumscript.sh /usr/local/bin/gitconvit
```
## Usage
To generate and use a conventional commit message, run the script:
```bash
gitconvit
```
#### Script Workflow
1. Check if you are in a Git repository.
2. Detect the changed files.
3. Generate a commit message using the AI model.
4. Prompt you to confirm or regenerate the commit message.
5. Commit and push the changes to your repository.
### Detailed Steps
1. **Check if the Current Directory is a Git Repository**:
* The script ensures that the current directory is a Git repository before proceeding.
2. **Get Changed Files and Diffs**:
* The script uses `git diff --name-only` and `git diff` to get the list of changed files and their diffs.
3. **Generate Commit Message**:
* The script sends the changes to the AI model `llama3:latest` from `ollama` to generate a conventional commit message with an appropriate emoji.
* The commit message follows the format specified by conventional commits.
4. **Interactive Confirmation**:
* The script prompts you to either accept the generated commit message, regenerate it by pressing `r`, or abort by pressing `CTRL+C`.
5. **Commit and Push**:
* Once confirmed, the script adds all changes, commits them with the generated message, and pushes to the current branch.
## Example
Here is an example of how the commit message might look:
```text
📝 docs(README.md): added installation method with docker
```
### Todo
- [ ] Add a flag to select different ollama model than llama3
- [ ] Add a flag to select different ai model than ollama like openai models or google gemeni or grok
- [ ] Add a tutorial GIF in the readme file
- [ ] Add an (add) option that will list all files changed and user can multiselect files with gum/fzf/case
- [ ] Make the emojis optional
- [X] Add an option to edit the AI output
- [X] Add an option to pass a prompt to the initial main prompt
- [X] Add an option to create the convential commit manually by taking three inputs the category, the context, and a short description
- [X] Add an option to recreate the AI output
## Contributing
Contributions are welcome! Please fork the repository and create a pull request with your changes.
## License
This project is licensed under the MIT License. See the LICENSE file for details.
## Contact
For any inquiries or support, please contact [Osama Yusuf](https://github.com/Osama-Yusuf).