ClawHub
Back to skill

Security audit

台灣房價估價分析 (Property Valuation)

Security checks across malware telemetry and agentic risk

Overview

This skill is a local Taiwan property valuation helper; its only notable issue is a disclosed optional analytics prompt that users should understand before using.

Before installing, be aware that valuation inputs may contain sensitive addresses and property details, so keep input and generated report files somewhere appropriate. The valuation script appears local-only. The analytics and ClawHub install-count references are disclosed but not necessary for valuation, so only run the analytics command or feedback channels if you are comfortable with that activity being recorded or shared.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Description-Behavior Mismatch

Low
Confidence
93% confidence
Finding
The skill is presented as a property valuation tool, but it also instructs users to run a separate analytics command that records usage. This introduces data collection behavior outside the core valuation purpose, creating a risk of undisclosed telemetry or metadata leakage, especially if users run it without informed consent.

Context-Inappropriate Capability

Low
Confidence
94% confidence
Finding
The documented usage analytics capability is not necessary to perform property valuation and therefore expands the skill's behavior beyond its stated purpose. Even if limited to usage counts, this can normalize unnecessary tracking and may expose user activity patterns, installation metadata, or other environment details if the analytics script is more capable than described.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.