Security audit
Security checks for vulnerabilities and agentic risk
This is a prompt-structuring skill that stays within text generation and does not request tools, credentials, persistence, or external actions.
Install this if you want help turning rough ideas into structured AI briefs and prompts. Expect it to ask clarifying questions and suggest downstream tools, but based on the artifact it should not perform external actions or change files by itself.
The display name and summary are presented only in Chinese, and the document consistently frames the skill in Chinese without any indication that users may choose another language. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale restriction is justified, which is not documented here.
The trigger condition is deliberately broad and based on fuzzy user intent rather than clear boundaries, so it can activate on many ordinary planning or brainstorming requests. That creates skill-selection ambiguity and increases the chance this skill intercepts requests that should go to a more specific or safer downstream skill, potentially degrading task routing and causing unintended prompt construction around sensitive requests.
