ClawHub
Back to skill

Security audit

综合美学设计师

Security checks across malware telemetry and agentic risk

Overview

This is a Chinese-oriented visual design helper with some quality and routing caveats, but no evidence of hidden access, persistence, or harmful behavior.

Install only if you want a Chinese-first visual design and art-direction assistant. Treat its critique helper as generic inspiration rather than grounded analysis unless the publisher updates it to actually evaluate the supplied design, and consider narrowing activation triggers or adding a language preference rule after installation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
98% confidence
Finding
The function claims to critique a supplied design description but never reads or analyzes the input, instead returning randomized canned output. In an agent skill, this is dangerous because downstream users or systems may rely on fabricated analysis as if it were grounded in the provided design, leading to deceptive or low-integrity results.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The documentation states that the tool evaluates a given design description, but the implementation does not do so. This mismatch is a security-relevant integrity issue because it can mislead operators into trusting the output as analysis of user input when it is actually unrelated random text.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger list includes very generic terms such as 'design' and 'aesthetic', which are likely to match many ordinary user requests and cause unintended activation. Over-broad activation increases prompt-scope collisions, making it easier for this skill to override or interfere with unrelated tasks and potentially expose users to irrelevant instructions or outputs.

Natural-Language Policy Violations

Medium
Confidence
84% confidence
Finding
The skill content is primarily written in Chinese and presents Chinese as the default interaction/output mode without an explicit user-language check. This can cause unintended language switching, reduce user comprehension, and make it harder for users or reviewers to notice problematic behavior or incorrect outputs.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.