ClawHub
Back to skill

Security audit

Mac System Stat

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed local Mac system-stat reporter, with no evidence of network exfiltration, credential access, destructive actions, or hidden persistence.

Install this only if you are comfortable letting the skill read and print local Mac system telemetry, including hostname, top process names, and GPU/driver details. Expect local Swift helper compilation on first use if Command Line Tools are available.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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)

subprocess module call

Medium
Category
Dangerous Code Execution
Content
def run_shell(cmd: str, timeout: int = 20) -> str:
    try:
        p = subprocess.run(cmd, shell=True, capture_output=True, text=True, timeout=timeout, executable=ZSH)
    except Exception as e:
        return f"ERROR: {e}"
    return (p.stdout or p.stderr or "").strip()
Confidence
97% confidence
Finding
p = subprocess.run(cmd, shell=True, capture_output=True, text=True, timeout=timeout, executable=ZSH)

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The helper enumerates GPU IOKit registry entries and emits detailed hardware and driver-related metadata, including registry IDs, model information, plugin/class names, and performance statistics, directly to stdout without any visible consent, minimization, or access control in the code. While this is non-privileged collection, it can still expose device fingerprinting data and low-level system details that may be sensitive in an agent context, especially if forwarded to logs, telemetry, or remote consumers.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.