Perfect your pull request and disclose you’ve used AI [AI Skills Series - Week Two]

[queenofcorgis]

Welcome to part two of our AI Skills Series. This week we’ll be writing a pull request commit message that discloses how you’ve used AI. As a reminder, those who participate will be entered to win one of 10 GitHub Shop vouchers.

Challenge

Pick one of the scenarios in the comments and write a commit message for it!

Scenario one: You’re writing a new unit test and are unsure if you’ve covered all edge cases. You ask GitHub Copilot for advice on potential edge cases to test. You select one suggestion and write the test logic yourself.

Scenario two: A function in the codebase is slow and needs optimization. Maintenance has requested more efficient code. You use GitHub Copilot to generate an optimized version of the function based on your description. You then review, adapt variable names, add comments, and thoroughly test the output.

Scenario three: The project is adding French language support. There are dozens of new UI strings that need translation. You use AI to translate all new strings, then work with a fluent French speaker in the community to review and correct the translations before submitting.

Scenario four: The maintainers invite proposals for a new plugin system. You want your proposal to be thorough and well-organized. You use AI to brainstorm ideas, outline the proposal, and draft much of the initial text. You then refine the proposal, add project-specific context, and verify technical feasibility before submission.

Scenario five: A utility function is slow and hard to read. The maintainers suggest refactoring for efficiency and clarity. You describe the function’s purpose to an AI code generator, which outputs a new, optimized implementation. You run tests, review for correctness, and add inline comments to clarify logic.

Examples

Good AI Disclosure Examples:

• Implement real-time chat feature with WebSocket connection handling
  Co-developed with AI - Human architecture decisions, AI implementation assistance for socket management
• Add OAuth2 authentication endpoints with rate limiting and JWT refresh AI-generated boilerplate - Used GPT-4 for initial implementation, customized security middleware
• Optimize API response caching and implement Redis connection pooling AI pair programming - Used GitHub Copilot for Redis setup, manually tuned connection parameters

Bad AI Disclosure Examples:

• Improve performance across multiple components AI assistance - Used AI for stuff
• Create innovative caching solution with advanced optimization techniques Novel implementation - Engineered from first principles using cutting-edge approaches (Actually: standard Redis implementation suggested by AI)
• Fix null pointer exception in user profile component Extensive AI session - Spent 4 hours with ChatGPT trying 23 different prompts, went through multiple conversations, had to explain the entire codebase context, tried Claude and GPT-4, finally got working solution on attempt 47

Resources and tools

• GitHub Copilot
• Helping other review your changes in a Pull Request

*No Purchase Necessary. Open only to Github community members 18+. Game ends 8/21/25. For details, see Official Rules.

[IK7D3]

Here's my submission for Scenario 5: Refactoring code.

Commit Message:

refactor: Simplify the user authentication logic

The previous implementation of the authenticateUser function was complex and hard to follow. This commit refactors the function for better readability and maintainability.

I used GitHub Copilot to brainstorm and suggest alternative approaches for structuring the conditional checks. The final implementation was written by me based on those suggestions, ensuring it aligns with our project's coding standards.

Co-authored-by: GitHub Copilot copilot@github.com

[stdavis]

Scenario 3 Commit Message

feat: add French language support

Added French translations for all UI strings. Used Google Gemini (via GitHub Copilot) 
for initial translations and then worked with a fluent, French-speaking human 
to refine and verify.

[Akash1134]

One more challenge to take you to the next level as a responsible AI contributor, come play code critic in our final challenge discussion.

[Rampyaaryans]

/assign

[Ubayed-Bin-Sufian]

Scenario 1: Add unit tests

Commit Message

test: add unit test for edge cases

This commit adds a test that covers a set of important edge conditions (empty input, null/undefined, invalid types, and boundary values), includes setup and teardown where needed, and asserts correct behavior and/or expected error handling. No production code changes.

Co-authored-by: GitHub Copilot copilot@github.com