Humanizer

Security checks across malware telemetry and agentic risk

Overview

This is a text-editing skill that is open about making AI-sounding writing more natural, with no code execution, network access, persistence, or hidden install behavior.

Safe to install from a technical security perspective. Use it on pasted text or clearly selected files, review edits before accepting them, and avoid using it where academic, workplace, legal, or platform rules require preserving the original authorial voice or disclosing AI assistance.

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 (1)

Natural-Language Policy Violations

Medium

Confidence: 90% confidence
Finding: The skill explicitly instructs the agent to inject first-person perspective, opinions, humor, and personality into rewritten text without requiring user consent. That can cause unintended impersonation or alteration of authorial voice, especially when editing professional, academic, legal, or organizational content where neutrality and attribution matter.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal