ClawHub
Back to skill

Security audit

Shuf Tool

Security checks across malware telemetry and agentic risk

Overview

This is a small local line-shuffling utility with a real stdin-handling bug, but no evidence of hidden exfiltration, persistence, credentials, network access, or destructive behavior.

Install only if you are comfortable with a simple local helper that can read files available to the agent. Until fixed, avoid piping sensitive text into it and do not rely on the documented -n, -i, -r, or -e options.

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)

Tp4

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding
The documented behavior says the skill shuffles input lines, but the analyzed behavior indicates it can instead read arbitrary filesystem paths and may treat stdin as a path rather than as text content. This mismatch is dangerous because users and downstream agents may pass sensitive filenames or untrusted input under the assumption of harmless text shuffling, enabling unintended file access and data disclosure outside the declared scope.

Description-Behavior Mismatch

Medium
Confidence
98% confidence
Finding
The script claims to shuffle text input, but when no command-line argument is provided it reads all of stdin and treats that data as a filesystem path, then opens that path. In an agent context this can cause unintended local file access or disclosure if upstream components pass text via stdin as expected, making the behavior dangerous and inconsistent with the advertised skill semantics.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.