pr-desc

What This Skill Does

The pr-desc skill is an intelligent utility designed to automate the often tedious task of drafting pull request descriptions. By leveraging advanced natural language processing, it inspects your local Git branch changes, identifies key modifications, and distills them into a clean, professional, and structured markdown format. This tool eliminates the need for manual write-ups, ensuring that your team members can quickly grasp the context, reasoning, and technical impact of your code changes without sifting through raw diffs.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/lxgicstudios/ai-pr-desc

Ensure you have Node.js 18 or higher installed on your machine. Additionally, you must set your OPENAI_API_KEY as an environment variable to authorize the generation of content. The tool performs best when run from the root of a initialized Git repository.

Use Cases

This skill is perfect for developers working in fast-paced teams who prioritize high-quality documentation but suffer from the overhead of manual maintenance. Use it for standard feature implementations, critical bug fixes, or minor refactoring tasks. It is particularly effective for larger PRs where manual categorization of "breaking changes" vs. "feature updates" can be time-consuming or error-prone.

Example Prompts

1. "Generate a detailed PR description for my current branch, focusing on the new authentication module and making sure to highlight the breaking changes in the API."
2. "Compare my feature branch 'refactor-dashboard' against 'main' and generate a PR summary that includes a section for testing instructions."
3. "Analyze the changes in the current branch and output a summary that includes a placeholder for visual screenshots of the UI updates."

Tips & Limitations

For best results, maintain clean, descriptive commit messages, as these provide context for the AI analysis. The skill relies on external OpenAI APIs, so ensure your connectivity is stable during execution. Note that the output is a draft; while highly accurate, it is always recommended to perform a quick review to ensure specific project-unique nuances are captured correctly. Always verify that your repository does not contain sensitive secrets before scanning, as the content is sent to the LLM for analysis.