Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese debate-roleplay skill with purpose-aligned web search and no hidden installation, persistence, or destructive behavior.
Install only if you want a Chinese-language structured debate assistant. Avoid entering sensitive personal, workplace, legal, medical, or private topics unless you are comfortable with the topic or derived search queries being sent to Tavily/web_search.
The description, trigger phrases, and required interaction examples are entirely in Chinese, and the skill instructs fixed Chinese reply formats without mentioning any user language choice. This can violate language or locale policy when a skill implicitly requires a specific language without opt-in or documented justification.
The read_when conditions are high-level and ambiguous, such as '需要 AI 模拟辩论场景', which can cause the skill to load in contexts where the user only wants discussion or analysis rather than this specific skill. Overbroad activation increases the chance of unrequested behavior and unnecessary external search exposure.
The trigger phrases are intentionally broad and explicitly use fuzzy matching, so normal conversational phrases like '来场辩论' may activate the skill when the user did not intend to enter a structured debate flow. This can cause unintended mode switching and, in this skill, may lead to unexpected use of web_search once a topic is supplied.
The skill states that Tavily search is preferred for evidence gathering but does not warn users that their debate topic may be sent to an external service. Debate topics may include sensitive personal, political, legal, health, or workplace scenarios, so undisclosed outbound transmission creates a privacy and consent risk.
No suspicious patterns detected.
