The Agent Skills Directory

[SAFE]: The skill operates purely as an advisory resource. It consists of markdown-based instructions and rule files that guide the agent's analysis process without involving remote script execution or external downloads.\n- [COMMAND_EXECUTION]: The workflow suggests using git diff --name-only HEAD~1 to identify changed files for review. This is a standard local development command used to provide context for the advisory task and is considered benign within this scope.\n- [DATA_EXFILTRATION]: There are no network operations or instructions to send data to external servers. The skill explicitly provides guidance on security best practices, such as warning against accepting secrets via command-line flags or environment variables in the tools being reviewed.

dx