Security audit
Security checks for vulnerabilities and agentic risk
This skill is a coherent UIBC medical-device judging aid with a small local checker script and no evidence of hidden access, persistence, exfiltration, or destructive behavior.
Install this if you need the UIBC MED-domain judging framework. Treat it as a specialized competition/governance aid, not as general medical, legal, regulatory, financial, or contract advice; keep prompts explicitly tied to TH-MED-001/002/003 adjudication when invoking it.
This markdown file contains user-facing operational instructions exclusively in Chinese, including the title, purpose, and invocation guidance. Under the language/locale policy rule, forcing a specific language without user opt-in or a documented justification can be a policy violation.
The description embeds a long list of broad trigger terms such as UIBC, medical devices, migration judgment, evidence package, and accountability, which can cause the skill to activate for loosely related prompts rather than only for the intended evaluation workflow. Over-broad invocation increases the chance of misrouting user requests into a specialized governance skill, creating unreliable outputs or inappropriate authority framing.
The read_when conditions rely on generic topic mentions and adjacent concepts like UIBC, medical-device teams, anti-fabrication guardrails, and design change, without defining firm boundaries. This can cause unintended invocation whenever a conversation merely references these topics, leading to scope creep and potentially authoritative-sounding judgments outside the intended competition-review context.
The usage guidance repeats the same broad trigger vocabulary but does not define scope limits, disambiguation rules, or negative examples. In practice, this can reinforce accidental activation and over-application of the skill's domain-specific judgment framework to unrelated requests, reducing safety and correctness.
