The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the GitHub CLI (gh) and Git to perform operations like viewing PR metadata, adding labels, and merging code. These are standard operations for development workflows.
[SAFE]: The skill implements logic to honor project-specific configuration (e.g., human_merge_required). It also includes the disable-model-invocation: true directive, which ensures the AI agent does not recursively process potentially malicious instructions embedded in pull request titles or comments, mitigating indirect prompt injection risks.

merge-pr