The Agent Skills Directory

[SAFE]: The skill utilizes legitimate command-line tools (git and gh) to perform pull request operations. These are standard for developer-oriented AI agent skills.
[SAFE]: Instructions include explicit safety rules that prevent dangerous or non-standard practices, such as creating PRs from the main branch or force-pushing without lease checks.
[SAFE]: Data handled by the skill (commits, diffs, and issue IDs) is processed within the scope of creating a pull request and is sent only to GitHub's official services via the CLI.
[SAFE]: No obfuscation, credential harvesting, or unauthorized remote code execution patterns were detected. The example issue ID 'SOYIO-20' aligns with the skill author's context.

create-pr