AI Code Review

• Automatically generate an AI code review on every PR.
• Easily customize the AI prompt as needed.
• Works with both Azure OpenAI or OpenAI APIs.
• Add to any GitHub repo with only 3 files.
• View Sample AI Code Review

Solution Files

1. ai_code_review.yml
   GitHub Actions file triggered upon pull request.

2. ai_code_review.py
   Python script to generate AI code review.

3. requirements.txt
   Python package dependencies (only openai and requests).

How to Use

1. Open or create a GitHub repo.
2. Confirm or create the .github/workflows/ directory.
3. Add 3 files from this repo to your project.
4. Choose your AI provider:
   • Azure OpenAI is highly recommended for private and enterprise use.
   • OpenAI is for public repos or when you give training consent.
5. Update your GitHub repo settings:
   1. Browse to Repo > Settings > Security and variables > Actions
   2. Add action secrets:
      • OPENAI_API_KEY
      • AZURE_OPENAI_ENDPOINT
      • AZURE_OPENAI_API_KEY
      • AZURE_OPENAI_API_VERSION (optional, default 2024-02-15-preview)
      • OPENAI_MODEL_NAME (optional, default gpt-4, or Azure deployment name)
      • Learn more about encrypted secrets
6. Create new branch, push changes, and create new PR.
7. Improve your code based on the automated AI feedback.
   • Comment with AI code review posted on PR.
   • workflow.log created within each action run.