ClawHub
Back to skill

Security audit

Avoid AI Writing

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed writing-editing skill that rewrites text you provide and does not show hidden execution, credential access, data exfiltration, or background persistence.

Install this if you want an assistant to clean up drafts for AI-sounding patterns. Expect it to read and transform text you paste or files you explicitly point it at, so avoid feeding it drafts you do not want your assistant environment to process. Prefer the ClawHub install path or verify the GitHub repository before manually cloning it into a skills directory.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • 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 (1)

Session Persistence

Medium
Category
Rogue Agent
Content
A portable writing skill for [Claude Code](https://docs.anthropic.com/en/docs/claude-code), [OpenClaw](https://github.com/openclaw/openclaw), and any [agentskills.io](https://agentskills.io)-compatible agent. Audits and rewrites content to remove AI writing patterns ("AI-isms").

Paste any text or run against a file, get back a clean version with every AI tell identified, fixed, and explained. A built-in second pass re-reads the rewrite to catch patterns that survived the first edit.

## Quick demo
Confidence
60% confidence
Finding
write to catch patterns that survived the first edit. ## Quick demo **Input:** > Certainly! Acme Analytics, a vibrant startup nestled in the heart of Boulder's thriving tech ecosystem, has secured $

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.