ClawHub

Writing Style Humanizer English

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only writing editor, but it is explicitly built to make AI-written text look unaided and human-written, which needs careful review before use.

Install only if you want a rewriting tool for text you are allowed to revise. Do not use it to bypass academic, workplace, publishing, legal, or platform rules that require disclosing AI assistance. Review outputs for invented first-person claims, unsupported opinions, or changes to the author's actual meaning.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Ssd 1

Medium
Confidence
97% confidence
Finding
The instruction to make text 'not obviously AI generated' explicitly frames the skill as helping conceal AI authorship and evade AI-origin detection. That can be misused for academic dishonesty, policy evasion, deceptive content laundering, or bypassing review processes that rely on disclosure of AI assistance.

Ssd 4

Medium
Confidence
98% confidence
Finding
The two-step workflow first audits text for AI-detection signals and then iteratively revises to remove those signals, which operationalizes evasion rather than neutral editing. This increases the skill's usefulness for systematically laundering AI-generated content into forms less likely to be flagged by human or automated reviewers.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal