Security audit
Security checks for vulnerabilities and agentic risk
This is a straightforward Chinese text-rewriting skill with no hidden execution, network access, persistence, or credential handling in the inspected artifacts.
Install this if you want a Chinese-focused rewriting assistant. Use explicit prompts or delimited text blocks when invoking it, especially if you only want analysis rather than automatic rewriting.
The README allows invocation not only via the explicit /humanizer-cn command but also by asking users to 'paste text and state your intent directly'. That broad trigger can cause the skill to activate in loosely matched conversational contexts, increasing the chance of unintended prompt interception or modification of unrelated Chinese text during normal chats.
The description says the skill is specifically for identifying and removing AI-writing markers from Chinese text, which imposes a language constraint. In this file, the skill does not state that users can choose another language or that the Chinese-only behavior is an explicit opt-in, creating a potential language/locale policy issue.
The README instructs users they can invoke the skill with a broad natural-language phrase (“去除 AI 痕迹”), which increases the chance of accidental or overly eager activation in unrelated contexts. In agent environments, ambiguous triggers can cause unintended transformation of user text, including rewriting content the user did not clearly consent to modify.
No suspicious patterns detected.
