🤖 GPT-PR is a command-line tool that leverages the power of OpenAI's GPT-3 and GPT-4 models to automatically generate pull request (PR) details, including the title, branch name, and description, based on a given Git diff.
All PRs of this repository have been created with GPT-PR.
- Generate PR title, branch name, and description using GPT-3 and GPT-4 models
- Automatically create and push a new branch with the generated details
- Create a pull request on GitHub using the generated details
- Clone the repository:
git clone https://github.com/javiln8/gpt-pr.git
- Build the project:
cd gpt-pr
go build -o gpt-pr
- Add the gpt-pr binary to your system's PATH.
Set up the GPT_API_KEY
environment variable with your OpenAI API key:
export GPT_API_KEY=<your_openai_api_key>
Set up the GITHUB_TOKEN
environment variable with your GitHub Personal Access Token:
export GITHUB_TOKEN=<your_github_personal_access_token>
To use GPT-PR, you can run the following commands:
Usage:
gpt-pr generate [flags]
Flags:
-a, --add-attribution Add an attribution message at the end of the PR summary
-g, --gpt-version int Choose the GPT version (3 or 4), default is 3 (default 3)
-h, --help help for generate
-v, --verbose Show the ChatGPT response
Although GPT-PR automates the PR creation process, it might be less reliable due to GPT-3's limitations. The tool makes multiple calls to ChatGPT for better results, but the generated PR details may still vary in accuracy. However, the tool serves as a valuable starting point for streamlining the PR creation process and can be further enhanced with more sophisticated prompts or by utilizing newer AI models.
To get API keys for GPT-3 or GPT-4, you need to sign up for an account with OpenAI. Visit the OpenAI website to create an account and obtain the necessary API keys (link).
This project was built with the help of OpenAI's GPT-4, which contributed to the code generation and decision-making processes.