ClawHub
Back to skill

Security audit

molt-board-art-free

Security checks across malware telemetry and agentic risk

Overview

This pixel-art board skill has a coherent core purpose, but it also asks for broad command and file authority and says it can activate for unrelated Operations work.

Review before installing. Use this only for the artboard workflow you expect: registering a bot, checking cooldown, placing pixels, and viewing canvas regions. Do not let it activate for generic Operations, batch processing, format conversion, or broad local troubleshooting unless you have separately inspected the actual script and trust its behavior. Protect the API key and credentials file, and confirm where artboard.sh comes from before running it.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (4)

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The skill is presented as a narrowly scoped pixel-art canvas tool, but these sections broaden it into generic operations, file handling, and command troubleshooting. That mismatch can cause an agent or user to grant broader trust and invoke exec/file capabilities outside the intended domain, increasing the chance of misuse or unsafe command execution.

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The manifest and description mix a simple artboard skill with broad automation, batch processing, format conversion, and command-execution framing. In a skill that exposes read/write/exec tools, this kind of scope inflation is dangerous because it can induce overly broad activation and trust for actions unrelated to the stated purpose.

Context-Inappropriate Capability

Low
Confidence
85% confidence
Finding
Documenting generic command execution and file-oriented troubleshooting exceeds what a basic pixel-art skill should require. Even without explicit exploit code, this normalizes broader shell/file usage and may lead an agent to perform unnecessary or risky local operations under the skill's authority.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger condition 'when the user needs Operations-related tasks' is far broader than the skill's stated artboard purpose. Overbroad auto-activation can cause the skill to engage in unrelated contexts, exposing exec/read/write capabilities and increasing the risk of unintended command execution or data access.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.