Security audit
Security checks for vulnerabilities and agentic risk
This is a plain Markdown medical-report explanation skill with disclosed limits and no executable behavior, though users should be careful because it handles sensitive health information.
Install only if you are comfortable sharing health-report details in the conversation. Treat outputs as preparation for a doctor visit, not as a diagnosis or medication plan, and be explicit when invoking the skill so generic health discussion is not accidentally interpreted as a report-analysis request.
The trigger keywords are broad, generic phrases like '指标偏高' and '体检异常' that are likely to appear in ordinary medical discussions. This can cause unintended invocation of the skill in unrelated conversations involving sensitive health content, increasing the chance of overreach or accidental processing of medical information.
The natural-language examples are open-ended and conversational, which may encourage activation from loosely related follow-up messages such as '按上面的结果改成更简短的版本'. In a medical context, unintended continuation is risky because it can keep the skill engaged with health-related content without a fresh, explicit user request.
No suspicious patterns detected.
