AI灾害应急助手

• Outcome: Pass
• Latest audit: Jun 14, 2026 at 11:55 AM UTC
• Version:
  1.0.0

Overview

This appears to be a Chinese emergency-preparedness guidance skill, with safety caveats to consider but no evidence of hidden access, persistence, data theft, or destructive behavior.

Install only if you can read Chinese comfortably and treat the content as general preparedness guidance, not a substitute for local emergency services, official alerts, certified first-aid training, or medical/rescue professionals.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

High

• Confidence: 97% confidence
• Finding: The skill provides extensive emergency, disaster-response, and first-aid instructions, including CPR, AED, airway obstruction, bleeding control, fractures, burns, and disaster evacuation, but it does not prominently warn that the guidance may be incomplete, situation-dependent, or not a substitute for local emergency services and trained responders. In a life-threatening context, users may rely on simplified or generic instructions instead of immediately contacting emergency services, which can lead to delayed care, inappropriate actions, or physical harm.

Natural-Language Policy Violations

Medium

• Confidence: 83% confidence
• Finding: The skill is presented entirely in Chinese without offering a language option or documenting that it is intentionally restricted to Chinese-speaking users. In emergency and first-aid scenarios, language mismatch can cause dangerous misunderstanding or make urgent instructions unusable for non-Chinese readers who encounter the skill.

VirusTotal

64/64 vendors flagged this skill as clean.