ClawHub
Back to skill

Security audit

旅行保险聪明买

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a travel-insurance advice tool, but it declares a proxy token despite saying it needs no API key.

Review before installing. The main issue is not detected exfiltration, but the mismatch between a no-API-key knowledge-base claim and a declared PROXY_TOKEN. Install only if you are comfortable with that declared environment access, or ask the publisher to remove or clearly document it.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Intent-Code Divergence

Medium
Confidence
89% confidence
Finding
The skill documentation says it is a 'pure knowledge-base skill' with 'no API Key required', yet the manifest declares a PROXY_TOKEN environment dependency. This mismatch can mislead users and operators about the skill's true trust boundary, increasing the risk of unnecessary secret exposure or accidental token provisioning to a skill that appears not to need it.

Vague Triggers

Medium
Confidence
74% confidence
Finding
The activation description uses broad trigger phrases around travel insurance and trip protection without clear exclusion conditions, so the skill may activate in loosely related conversations. Over-broad activation can cause unintended context capture, irrelevant recommendations, and routing of user queries into a skill that includes purchase-link guidance, which is risky in financial or insurance-adjacent contexts.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.