The Agent Skills Directory

[DATA_EXPOSURE]: Technical reference documentation in 'SKILL.md' and 'references/permissions-reference.md' mentions sensitive paths such as '/etc/shadow', '.env' files, and SSH credentials. These occurrences are used solely as examples to demonstrate the configuration of security denial rules and do not constitute an actual data exposure risk within the skill itself.- [DYNAMIC_EXECUTION]: The documentation references 'PreToolUse' hooks, which allow for custom logic to be executed before tool calls. This is presented as a legitimate platform mechanism for extending the security policy engine rather than a vector for unauthorized code execution.- [COMMAND_EXECUTION]: The skill provides examples of Bash commands (e.g., 'npm run build', 'git commit') to illustrate how to define permission allowlists. These are documented for user configuration purposes and are not executed by the skill automatically.

review-permissions