Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent low-carb diet advisor, but it asks for sensitive health details and gives condition-specific diet or fasting guidance without enough privacy and clinical-safety boundaries.
Review carefully before installing. Treat it as general lifestyle information, avoid sharing unnecessary identifiers or sensitive medical records/photos, and consult a licensed clinician before using it for diabetes, pregnancy, kidney/liver/heart disease, eating-disorder history, acute symptoms, or medication-related decisions.
The manifest description and the entire instruction set are written as a Chinese-only consultation skill, and the reply style does not indicate that users may choose another language. This creates a locale/language policy issue because the skill appears to mandate one language without opt-in or alternative selection.
The skill claims it does not diagnose, but it gives condition-specific dietary regimens, contraindication handling, fasting guidance, and a default meal-scoring framework that can function as individualized medical advice. In a medical/nutrition context, users may rely on this guidance for diabetes, insulin resistance, or medication-sensitive conditions, creating a real risk of harm despite the disclaimer.
The skill explicitly asks for health status, medications, symptoms, and lab-like metrics, and also invites meal photo analysis, all of which may contain sensitive health information. Without an explicit privacy notice, minimization guidance, and handling boundaries, users can be induced to overshare regulated or highly sensitive personal data.
No suspicious patterns detected.
