网红海岛

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:27 PM UTC
• Version:
  0.1.0

Overview

This is a simple travel-planning skill with some topic ambiguity, but it does not request code execution, credentials, persistence, or sensitive access.

Safe to install from a security perspective, but treat its recommendations as rough planning help. Because the skill mixes island and hiking-route concepts, verify destination type, access, ferry or transit options, weather, closures, permits, prices, and local safety rules before relying on it for travel.

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 (1)

Description-Behavior Mismatch

Medium

• Confidence: 90% confidence
• Finding: The skill claims to provide recommendations and guides for internet-famous islands, but the body content describes hiking/trail filtering and urban walking-route fields instead. This semantic mismatch can cause the agent to invoke the wrong skill for user travel queries, producing misleading safety, routing, or planning advice in a domain different from what the user requested. In a travel context, inaccurate scope and mislabeled capabilities increase the chance of bad recommendations and user harm from relying on irrelevant conditions or route assumptions.

VirusTotal

63/63 vendors flagged this skill as clean.