This GitHub Action uses patchwork-cli to automatically summarize and comment on PRs in your repository.

• Automatically reviews and summarizes pull requests
• Generates improvement suggestions for the PR (optional)
• Configurable summary length (long, short, none)
• Supports various LLM providers (OpenAI, local models, or custom endpoints)
• Automatic PR detection in GitHub Actions workflows

name: PR Review
on:
  pull_request:
    types: [opened]

jobs:
  review:
    runs-on: ubuntu-latest
    steps:
      - uses: patched-codes/[email protected]
        with:
          patched_api_key: ${{ secrets.PATCHED_API_KEY }}
          # PR URL is automatically detected in pull_request events

One of the following must be provided:

• patched_api_key: Patched API key for the LLM (Get an API key)
• openai_api_key: OpenAI API key for the LLM (Get an API key)

• pr_url: URL of the pull request to review (default: automatically detected in pull_request events)
• github_token: GitHub token for creating PR comments (default: Automatically set by GitHub)
• model: LLM model to use
• client_base_url: Base URL for the LLM API
• diff_suggestion: Whether to generate suggestions to improve the PR
• diff_summary: Length of the summary (long, short, none)
• model_temperature: Temperature parameter for the LLM
• model_top_p: Top-p parameter for the LLM
• model_max_tokens: Maximum tokens for the LLM response

- uses: patched-codes/[email protected]
  with:
    patched_api_key: ${{ secrets.PATCHED_API_KEY }}

- uses: patched-codes/[email protected]
  with:
    patched_api_key: ${{ secrets.PATCHED_API_KEY }}
    pr_url: "https://github.com/user/repo/pull/123"
    diff_suggestion: true
    diff_summary: "long"

- uses: patched-codes/[email protected]
  with:
    openai_api_key: ${{ secrets.OPENAI_API_KEY }}
    model: "gpt-4"
    model_temperature: 0.2
    model_top_p: 0.95
    model_max_tokens: 2000

Here are some example comments generated with the PRReview action:

• patched-codes/patchwork#59 (comment)
• patched-codes/patchwork#56 (comment)

MIT

Contributions are welcome! Please feel free to submit a Pull Request.