The Agent Skills Directory

[SAFE]: The skill is a documentation and utility guide for establishing safety guardrails. It instructs the agent on how to create rules that monitor for and block potentially harmful actions like destructive file deletion (rm -rf) or the inclusion of secrets in configuration files (.env).
[COMMAND_EXECUTION]: Provides instructions for local regex validation using Python. This is a standard development practice for testing the logic of the rules being created and does not involve remote code execution, obfuscation, or suspicious network activity.

writing-rules