The Agent Skills Directory

[SAFE]: No malicious patterns, such as prompt injection or safety bypasses, were detected in the skill instructions. The instructions are aligned with the stated purpose of editing text for style and tone. \n- [SAFE]: The skill requests standard file-system tool permissions (Read, Write, Edit, Grep, Glob) and user interaction (AskUserQuestion), which are appropriate for a text editing tool. No network or administrative tools are requested. \n- [SAFE]: No sensitive data exposure, hardcoded credentials, or obfuscated content was identified. \n- [SAFE]: Although the README provides installation instructions involving a third-party GitHub repository, these are directed at the user and do not involve the skill performing automated or hidden remote code execution.

humanizer