The Agent Skills Directory

[SAFE]: No security issues or malicious patterns were detected. The skill is designed for local operations and demonstrates defensive coding practices, such as environment poisoning protection for Python-based configuration reading.- [COMMAND_EXECUTION]: The skill executes internal Node.js and Bash scripts to manage log lifecycle and resolve workspace paths. These executions are scoped to the skill's own assets and include significant hardening against improper path resolution or shell-level interpretation errors.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data (such as commit history and session logs) to identify improvements. This potential attack surface is addressed through a strict human-in-the-loop gate; the skill creates staged copies of suggested updates but nunca modifies live skill files automatically.

observe-work