ClawHub

Zhua Metacognition

Security checks across malware telemetry and agentic risk

Overview

This skill appears to be a harmless metacognition prompt package, though it is incomplete and mostly Chinese-language content.

Install only if you want a lightweight self-reflection checklist. Treat the documented command examples as nonfunctional unless the missing scripts are later supplied, and avoid fetching substitute scripts from untrusted sources.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding
This is a mismatch because the declared purpose describes a substantive metacognitive capability, but the actual code does not implement that purpose at all. The script merely prints an example message and includes comments indicating it is a placeholder. There is no hidden behavior or unrelated resource access, but the primary purpose of the code differs materially from the description because no described functionality is present.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The manifest description and main skill content are written entirely in Chinese, and there is no indication that users may choose another language or that the skill is intentionally limited to a Chinese-only context. Under the policy, forcing a specific language without opt-in is a natural-language policy violation.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal