Security audit
Security checks for vulnerabilities and agentic risk
This is a text-rewriting skill with ordinary document-editing access and no hidden code, persistence, or data-export behavior.
Before installing, understand that this skill may rewrite tone and style aggressively, including adding more personal voice. Use it on selected or reviewed text where you are comfortable accepting or rejecting edits, especially for formal, legal, academic, or brand-sensitive writing.
The phrase "Or invoke directly when editing documents" does not specify what exact trigger, context, or constraints cause the skill to activate. Because this is a markdown file and the usage description lacks specific trigger phrases or exclusion conditions, it creates ambiguity about when the skill should run versus when it should not.
This markdown/manifest file says to use the skill when "editing or reviewing text to make it sound more natural and human-written," which is a very broad condition that overlaps with common writing assistance tasks. It does not provide explicit trigger phrases, scope limits, or negative examples clarifying when the skill should or should not activate.
The usage section says "Ask your agent to humanize text" and gives a single example, but it does not define the full allowed trigger set or boundaries for invocation. In a markdown skill description, this can be too open-ended because it provides no negative examples or explicit limits on whether the skill applies to whole documents, selections, or arbitrary conversational requests.
No suspicious patterns detected.
