ClawHub
Back to skill

Security audit

Linux

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed Linux administration skill that keeps local notes and gives high-impact host guidance, but I found no hidden execution, exfiltration, or purpose-mismatched behavior.

Install only if you want an agent to keep local operational notes about your Linux hosts and to assist with powerful admin workflows. Review destructive, remote-access, storage, account, and firewall changes before allowing them, and treat the flagged chown guidance as needing correction in a future update.

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
91% confidence
Finding
The documentation states that `chown -R` follows symlinks outside the target, which is generally inaccurate for typical GNU/Linux `chown` defaults and can mislead operators about how ownership changes behave. In a hardening/debugging skill, incorrect filesystem-safety guidance can cause admins to choose the wrong mitigation or avoid valid commands during incident response.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
memory-template.md:56