Back to skill

Security audit

Sleep Quality Advisor

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple Chinese-language sleep-habit advisor with no evidence of hidden access, persistence, exfiltration, or destructive behavior.

Install only if you want a Chinese-language, informational sleep-habit advisor. Treat its score and suggestions as lifestyle guidance, not medical diagnosis or treatment, and seek professional care for persistent, severe, or concerning sleep symptoms.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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
Findings (9)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill is framed as assessing sleep quality, scoring symptoms, and giving recommendations and 'medicalAdvice' without any disclaimer, limitation, or escalation guidance. In a health-related context, this can mislead users into treating the output as medical advice, potentially delaying professional care for serious causes of insomnia or sleep disturbance.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The trigger phrases are very broad and can match common sleep-related queries without clear boundaries, increasing the chance the skill activates outside its intended context. Unintended activation is especially concerning here because the skill proceeds to evaluate symptoms and provide quasi-medical guidance, which could cause users to rely on unsuitable advice.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill logic detects Chinese query terms such as '2点', '手机', and '很累', and the default input/output strings are also entirely in Chinese. This creates a language-specific behavior with no user opt-in or documented locale justification, which matches the language/locale policy violation criteria.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill generates a quantified sleep quality score, labels the user's condition, and gives a consultation recommendation, yet the disclaimer frames the output as merely general lifestyle advice. This can mislead users into giving the tool undue medical credibility or delaying appropriate care because the assessment appears clinical despite lacking validation or safety guardrails.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The code detects only Chinese trigger terms such as '2点', '手机', '很累', and returns all recommendations and labels in Chinese. This imposes a specific language/locale on all users without any opt-in or documented justification, which is a natural-language policy violation.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/test-handler.py (reported line 3)May include surrounding context.

python
#!/usr/bin/env python3
import json, subprocess
out=subprocess.check_output(['python3','scripts/handler.py','我最近总是凌晨2点才睡着,早上7点起床,很累'], text=True)
data=json.loads(out)
assert 'sleepAssessment' in data
assert 'medicalAdvice' in data

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The description 'auto-generated skeleton' is too vague to communicate the skill's purpose, boundaries, or expected behavior. This can cause unintended activation, misuse, or inadequate review because users and defenders cannot easily tell what the skill is supposed to do from its manifest.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

All headings, triggers, and workflow text are written in Chinese, which effectively forces a specific language experience. The file does not indicate that this is a region-specific skill or that users may opt into another language.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The test invokes the handler using a hardcoded Chinese prompt string, which reflects a language-specific interaction embedded in the file. Because there is no indication that language is configurable or user-selected, this may violate a policy requiring language/locale choice.

Content

No source excerpt is available for this finding.