ClawHub

Humanizer 1

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only writing editor that openly rewrites text to sound less AI-generated; the main risk is misuse to hide AI assistance, not hidden code or data access.

Install only if you want an agent to rewrite selected text or documents for style. Review changes before saving or publishing, especially for academic, legal, compliance, professional, or third-party-authored content, and do not use it to bypass disclosure or authorship rules.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The skill description is broad enough to trigger during ordinary editing or review work, not just clearly scoped stylistic cleanup. Because it is framed as generally making text sound more natural and human-written, it can be invoked in many contexts where disguising AI involvement or materially altering tone may be inappropriate. In this context, the broad trigger increases the chance of misuse rather than directly causing harm by itself.

Natural-Language Policy Violations

High
Confidence
97% confidence
Finding
The skill explicitly instructs rewriting text to remove indicators of AI generation and make it appear human-written. That directly enables deceptive authorship presentation in settings like schoolwork, journalism, reviews, applications, compliance writing, or social engineering, where the fact that content was AI-assisted may be material. The rest of the skill reinforces this by telling the model to 'add soul' and human signals, which makes the concealment objective more deliberate and more dangerous.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal