Undeclared Tool Scope
Medium
- Category
- MCP Least Privilege
- Confidence
- 70% confidence
- Finding
Without declared permissions the skill's intent is opaque and cannot be validated.
- Content
Security audit
Sleep Wind Down Coach
Security checks for vulnerabilities and agentic risk
Overview
The skill is mostly a simple sleep-routine helper, but its examples ask for diagnosis and mention a sleep aid despite saying it will not give medical advice.
Review before installing. Use this only for general sleep-hygiene routines, not for diagnosing insomnia or choosing supplements or sleep aids. The publisher should remove diagnostic framing and melatonin recommendations, and add clear guidance to seek qualified medical advice for persistent insomnia, severe daytime impairment, medication questions, or suspected medical issues.
Vulnerability Patterns
- Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
- Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
- Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
- Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
- Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
- MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 98% confidence
- Finding
The skill explicitly says it is not for diagnosis or treatment, yet the example scenario instructs the agent to 'Diagnose' sleep latency causes. That contradiction can push the agent into providing quasi-medical assessment without appropriate safeguards, which is risky in a health-related context because users may rely on incorrect or oversimplified guidance instead of proper clinical care.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
Lines L84-L86 present a usage scenario whose expected output is specifically in Chinese and framed for a Chinese locale ('中国打工人') without indicating that language choice should be based on user preference. This can violate language/locale policy because it normalizes forced localized responses rather than offering opt-in or matching the user's selected language explicitly.
- Content
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The expected output recommends melatonin even though the skill's safety section says not to present medication advice. Supplement or medication suggestions in a sleep-coaching skill can be inappropriate or unsafe because dosing, interactions, contraindications, and user-specific risks are not assessed.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The expected output for Scenario 1 is written as a mandated Chinese-language response pattern ('中国打工人睡前断电程序') rather than offering language choice. Under the policy, forcing a specific language or locale without user opt-in is a natural-language policy violation.
- Content
Static analysis
No suspicious patterns detected.