ClawHub
Back to skill

Security audit

Powershell Sandbox

Security checks across malware telemetry and agentic risk

Overview

This package claims to provide a safe PowerShell sandbox, but the actual sandbox implementation is missing while the docs encourage trusting it for untrusted scripts.

Review before installing. Do not rely on this package to sandbox untrusted PowerShell unless the missing src/sandbox.ps1 implementation and meaningful security tests are supplied and reviewed. If experimenting, run only trusted scripts in a VM or OS sandbox and require explicit approval before any PowerShell execution.

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 (1)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The document makes strong security assurances such as 'safe for untrusted scripts' and 'ready for production use' even though the same file explicitly lists unimplemented security tests and missing controls. This can mislead operators into deploying the sandbox in high-risk contexts with unjustified trust, increasing the chance that malicious scripts bypass incomplete protections.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.