Security audit
Security checks for vulnerabilities and agentic risk
This skill organizes user-provided health records locally with explicit privacy and consent boundaries.
Before installing, be aware that this skill is meant to process sensitive health records and create a persistent local folder of raw files, extracted notes, structured data, and summaries. Use it only for records you want organized this way, choose the storage location deliberately, and do not approve external OCR or API uploads unless you are comfortable sharing those health files with that provider.
The example triggers are very broad everyday phrases that could cause the skill to activate for loosely related requests without making its boundaries explicit. In a health-records context, overscoped activation is riskier because it may lead the agent to collect, organize, or infer highly sensitive medical information before the user has clearly consented to that specific workflow.
The sharing/positioning text describes building a personalized health knowledge system but does not warn that the workflow involves sensitive medical data, including reports, scans, and longitudinal records. In this context, missing notice and handling guidance increases the chance that users will provide protected or highly sensitive health information without informed awareness of privacy, retention, and exposure risks.
No suspicious patterns detected.
