ClawHub
Back to skill

Security audit

Curriculum Standard

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed document-generation workflow for Chinese vocational curriculum standards, with no hidden code or unrelated data access found.

Before installing, be aware that the skill may read uploaded training plans and related course documents, may download a public teaching-standard PDF after confirmation, and may create intermediate files/logs while generating the document. Review generated curriculum text carefully, especially any marked placeholders or inferred sections, before using it as an official standard.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The fallback section permits generating output with inferred structure or placeholder values when source information is missing, which conflicts with earlier requirements that prohibit guessing or using unconfirmed defaults. In a document-generation skill for formal educational standards, this can silently produce authoritative-looking but inaccurate compliance content, increasing the risk of downstream misuse or regulatory/academic errors.

Vague Triggers

Medium
Confidence
74% confidence
Finding
The skill uses a very broad trigger list and expansive 'must use' wording, which can cause the agent to invoke this workflow in loosely related contexts without strong boundary checks. Misrouting user requests into a rigid multi-phase document-generation process may lead to unnecessary file access, incorrect automation decisions, and production of inappropriate official-style documents from partial context.

VirusTotal

58/58 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.