ClawHub
Back to skill

Security audit

Visual Qa

Security checks across malware telemetry and agentic risk

Overview

This is a coherent visual regression testing skill, with expected local server execution and screenshot file persistence that users should handle carefully.

Install dependencies in a virtual environment or CI image you control. Only use --server or a config server value you trust, especially in unfamiliar repositories, because it runs with your local user permissions. Treat .visual-qa screenshots and diffs as potentially sensitive and review baseline deletion before replacing approved references.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

subprocess module call

Medium
Category
Dangerous Code Execution
Content
env = os.environ.copy()
    env['PORT'] = str(port)
    
    process = subprocess.Popen(
        command,
        shell=True,
        stdout=subprocess.PIPE,
Confidence
95% confidence
Finding
The script passes a user-controlled string from --server/config directly to subprocess.Popen with shell=True, which allows shell metacharacters and command chaining. In a skill/agent context, this can lead to arbitrary command execution on the host if untrusted input reaches that parameter.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The documentation instructs users to delete the entire baseline directory and replace it with current screenshots, but it does not warn that this permanently destroys the historical reference set. In a visual QA workflow, this can cause accidental loss of trusted baselines and make it easier to bless unintended or compromised UI changes without proper review.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The tool exposes arbitrary shell command execution via --server without any guardrails or warning, and the command is run through the shell. In an agent skill, this is more dangerous because higher-level orchestration may pass partially trusted parameters, turning a convenience feature into an RCE primitive.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.