ClawHub

AU Migration Visa File Manager

Security checks across malware telemetry and agentic risk

Overview

This is a static migration-practice drafting and checklist skill with a privacy inconsistency, but no hidden execution, persistence, credential use, or data exfiltration behavior.

Install only if you are comfortable using an AI assistant for sensitive migration-office drafting. Avoid entering full names, dates of birth, passport numbers, or other direct identifiers in chat unless your organization has approved that workflow; use initials or file references and insert legal names offline in final documents.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Confidence
96% confidence
Finding
The skill explicitly asks for a client's full name and country of origin in chat, which conflicts with its own privacy guardrail instructing users not to share identifying details. In a migration-services context, this increases unnecessary collection of sensitive personal data and raises the risk of privacy exposure through chat logs, transcript retention, or downstream model handling.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The template-drafting workflow asks for the client's full name and country of origin without opt-in, minimization, or a secure-channel warning, despite the skill later warning against sharing full names in chat. In a legal/migration setting, this can lead users to disclose personally identifiable and sensitive immigration-related information unnecessarily, increasing privacy and confidentiality risk.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal