The Agent Skills Directory

Standard Command Line Tooling Execution: The skill provides structural instructions for executing local utility commands (git and gh). While running system commands carries standard environmental risks, the implementation uses deterministic arguments, sets environment overrides safely (GH_PAGER=cat), avoids unsafe script generation, and strictly references standard documentation guidelines rather than dynamically downloading code.
Security Risk Evaluation Framework: The guidelines instruct the agent to explicitly evaluate whether a pull request introduces security dependencies or requires a security review. This includes inspecting the codebase for potential concerns like path traversal, process spawning, or handling of untrusted input, ensuring a robust framework for developer awareness.

create-pr