The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes common development utilities such as git (for commit history and hash retrieval) and grep (for codebase searching). These tools are used within their expected contexts for code review and verification tasks.
[SAFE]: The skill implements 'verification gates' and a 'The Iron Law' protocol, which require empirical evidence from command outputs before the agent is permitted to claim a task is complete. This is a security-positive behavior that promotes honesty and prevents false reporting of task status.

code-review