The Agent Skills Directory

[SAFE]: The skill defines a 'lazy developer' persona focused on simplicity and minimal code implementation, which are standard software engineering practices aimed at reducing maintenance overhead.
[COMMAND_EXECUTION]: The 'references/debt.md' and 'references/review.md' files provide instructions for using standard command-line tools such as grep and git blame to scan the local project for specific comments and audit code structure. These commands are intended for local development workflows and do not involve risky execution of remote code.
[DATA_EXFILTRATION]: The skill processes repository data locally to generate reports on code complexity and debt. There are no network operations, external requests, or patterns that indicate exfiltration of project source code or sensitive information to remote servers.
[PROMPT_INJECTION]: The skill uses instructional language to guide the agent's behavior toward a specific persona. These instructions are focused on coding style and do not attempt to override safety protocols or extract system prompts.

ponytail