AI-Generated Messages from Git Diffs and Logs

This script uses the OpenAI API to generate text based on the outputs of git diff and git log commands. It can be particularly useful for creating pull request descriptions, commit messages, or any other narrative that requires summarizing changes between different branches or commits in a Git repository.

Features

• Automated Message Generation: The script generates text based on Git diffs and logs using OpenAI's GPT models.
• Customizable Prompts: You can configure both the system and user prompts to guide the generated text.
• Flexible Configuration: Supports custom configuration through JSON files, allowing for different setups and use cases.
• Easy Configuration Management: The script can load configurations via command-line arguments or shorthand keywords, making it convenient to switch between different setups.

Prerequisites

• Python 3.x
• Git: Ensure that Git is installed and available in your system's PATH.
• OpenAI API Key: You need an OpenAI API key to use the GPT models.

Installation

Clone the repository and navigate to the directory:

git clone https://github.com/PxaMMaxP/ait
cd your-repository

Install the necessary Python packages:

pip install openai

Usage

Basic Usage

To generate a message based on the current branch's diff and log against the main branch, simply run:

python ait.py

This will use the default configuration file (ait.config.json) located in the current directory.

Using Custom Configurations

You can specify a different configuration file using the --config option:

python ait.py --config custom_config.json

Shorthand Configuration Selection

Alternatively, if you only provide a keyword, the script will look for a file named ait.<KEYWORD>.config.json in the current directory:

python ait.py --config custom

This command will look for a configuration file named ait.custom.config.

Fallback Configuration File search

As a fallback, if no configuration file is found, the script will search for it in the user's home directory under ~\.ait\*.

You can also provide a user wide configuration file by placing it in the ~\.ait\ directory.

Command-Line Options

• --config: Path to a JSON config file or a keyword to identify a specific config (ait.<KEYWORD>.config). Default is ait.config.json.
• --api_key: Your OpenAI API key. If not provided, the script looks for it in the config file.
• --diff_expression: The Git diff expression to use. Default is main....
• --log_expression: The Git log expression used to retrieve the commit history. Default is main....
• --system_prompt: The system prompt to set the behavior of the assistant.
• --prompt: The user prompt to guide the description creation.
• --model: The model to use, such as "gpt-4". You can change this depending on the available models in your API.
• --max_tokens: The maximum number of tokens (words or parts of words) in the API response. Default is 150.
• --temperature: Controls the randomness or creativity of the model's output. Lower values make the output more deterministic, while higher values make it more random. Default is 0.7.

Example Config Files

See the ait.commit.sample.config.json and ait.pull_request.sample.config.json files for examples of configuration files.

Error Handling

The script will raise an error if the API key is not provided either via the command-line argument or the configuration file.

License

This project is licensed under the MIT License. See the LICENSE file for more details.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.