Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese-language e-commerce research skill with no evidence of hidden execution, credential access, persistence, or destructive behavior.
Install this if you are comfortable with a Chinese-language e-commerce research assistant. Review outputs before using them for business decisions, and prefer explicit ecom-intel commands if you want to avoid accidental activation from broad research phrasing.
The skill metadata and description are entirely in Chinese and present the interaction model as Chinese-first without indicating language negotiation or fallback behavior. If enforced at runtime, this can override user language preference, degrade usability, and create safety issues where instructions, warnings, or consent cues are not understood by the user.
The trigger phrase 帮我深度研究一下 [研究对象] is overly broad and closely matches normal conversational input, so the skill may activate unintentionally during unrelated chats. This can cause prompt routing mistakes, unexpected data handling, or execution of the skill when the user did not explicitly opt in.
This shell script emits all user-facing messages in Chinese, including the title, test descriptions, and warning text. That imposes a specific language on users without opt-in, which matches the natural-language locale policy concern for forced language selection.
No suspicious patterns detected.
