gpt-pr-suggestions

What is gpt-pr-suggestions?

This project leverages the power of OpenAI's models to review and suggest improvements on code changes in a GitHub pull request.

When an user comments /suggest on a pull request, the model then returns suggestions and improvements, which the bot automatically comments on the pull request.

Installation

Before running the bot, you need to configure a few environment variables:

• GITHUB_TOKEN: Your GitHub API token.
• OPEN_AI_TOKEN: Your OpenAI API key to use the ChatGPT service.
• OPEN_AI_MODEL: OpenAI model, e.g., gpt-3.5-turbo.

1. Clone the repository: git clone https://github.com/RafaelBroseghini/gpt-pr-suggestions
2. Install the required dependencies: go mod download

Testing Locally

To test this bot locally, you'll need to set up the GITHUB_EVENT_PATH environment variable. This variable should point to a file containing a JSON representation of the GitHub event.

One of the easiest ways to get a realistic JSON event is to extract it from a real GitHub workflow.

To get this event payload:

1. Set up a workflow on your repository that gets triggered on comments to a PR.
2. In the workflow run, add a step that prints the contents of the GITHUB_EVENT_PATH file (cat $GITHUB_EVENT_PATH).
3. Run the workflow that includes the event you want to test (e.g., a comment on issue event).

Here's an example of such a step in a workflow:

on:
  issue_comment

jobs:
  show-event:    
    name: Print event
    runs-on: ubuntu-latest
    steps:
    - name: Dump GitHub event
      cat $GITHUB_EVENT_PATH

4. Go to a PR in the target repository and comment /suggest.
5. Check the workflow logs to find the printed JSON event data.
6. Copy the JSON data into a local file, and set the GITHUB_EVENT_PATH environment variable to the path of this file before running your bot locally.

For more detailed instructions, you can reference: Stack Overflow question.

Once the environment variables are set, you can run the bot with:

go run main.go