ClawHub
Back to skill

Security audit

Canvas Design

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only visual design skill that fits its stated purpose, though users should approve any external font downloads.

Install this if you want a skill that strongly steers an agent toward polished static artwork. Prefer bundled or local fonts, approve any external font downloads before they happen, and review generated .md, .png, or .pdf files before publishing or sharing them.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill explicitly authorizes downloading whatever fonts are needed, which expands behavior from local art generation into unconstrained network access. That creates risk of fetching untrusted remote content, leaking usage metadata, and enabling supply-chain or prompt-driven exfiltration behaviors unrelated to the user’s core request.

Vague Triggers

Medium
Confidence
82% confidence
Finding
The invocation description is very broad, so the skill may activate for many common art or design requests without strong narrowing conditions. Over-broad triggering can cause the agent to use this skill in situations where its extra behaviors, including file generation and asset acquisition instructions, are unnecessary or unexpected.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The markdown instructs the agent to download external fonts but provides no user-facing warning that network access may occur. This undermines transparency and informed consent, and in context is more dangerous because a design skill does not inherently require live network fetching when a local font directory is already referenced.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.