Security audit

顶级名医中医辨证Skill

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed local Chinese-medicine education skill, but it asks for sensitive health details and produces prescription-style herb guidance without clear consent or safety boundaries.

Review before installing. Use this only for educational TCM-style explanations, not diagnosis or treatment. Avoid sharing unnecessary personal health details, and do not rely on its herb formulas, dose amounts, child guidance, or pregnancy cautions without a licensed clinician.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The manifest advertises broad automatic completion of the full问诊→辨证→开方 workflow with '无需手动输入额外参数' and without clear trigger constraints or scope limits. In an agent ecosystem, this increases the chance of over-broad invocation in unrelated contexts and can cause unsolicited handling of sensitive medical information or generation of treatment advice without adequate user intent confirmation.

Missing User Warnings

High

Confidence: 97% confidence
Finding: The skill description says it will automatically ask for and process highly sensitive health data, including symptoms, menstrual history, pediatric development, tongue and pulse information, but it does not present an upfront privacy warning or consent gate. Because this is a medical skill, the lack of a clear notice about sensitive data collection, processing, and limitations materially raises privacy, compliance, and user-safety risk.

Static analysis

No suspicious patterns detected.