Ae4
Medium
- Category
- analysis-evasion
- Confidence
- 80% confidence
- Finding
Suspicious Unicode normalization or mixed-script content
- Content
Security audit
disease-cea-auto
Security checks for vulnerabilities and agentic risk
Overview
This skill provides disclosed drug cost-effectiveness analysis and does not show hidden data collection, persistence, or system-changing behavior.
Use this skill when a China-centered, bilingual pharmacoeconomic comparison is appropriate. Its outputs depend on live web sources and modeling assumptions, so users should verify citations, costs, treatment choices, and medical/economic conclusions before relying on them for policy, clinical, or purchasing decisions.
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
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The trigger list includes broad phrases such as cost-effectiveness analysis, multiple drugs, QALY, and treatment comparison, which can match ordinary user requests beyond the intended narrow workflow. Overbroad activation can cause the skill to intercept unrelated prompts, perform unintended web-driven analysis, and override user language or formatting expectations, increasing the chance of unsafe or confusing behavior in multi-skill environments.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 98% confidence
- Finding
The skill mandates bilingual Chinese-English output for every section without checking the user's preferred language. While not a code-execution risk, this can violate locale expectations, leak unnecessary translated content into downstream workflows, and reduce user control over output format, especially if another system assumes single-language responses for further processing.
- Content
Ae4
Medium
- Category
- analysis-evasion
- Confidence
- 80% confidence
- Finding
Suspicious Unicode normalization or mixed-script content
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The script instructs the AI to search for the latest China GDP per capita and uses a China national statistics source and Chinese WTP threshold as defaults. This imposes a specific national locale/policy context in the generated analysis without offering a user choice or clearly stating that the skill is restricted to China-specific evaluations.
- Content
Static analysis
No suspicious patterns detected.