AI Content Humanizer Pro
Security checks across malware telemetry and agentic risk
Overview
This is a disclosed text-rewriting skill with no executable code, hidden behavior, network access, or persistence, though users should review edits and avoid policy-sensitive misuse.
Install this only if you want an agent to rewrite user-provided text and selected files. Review diffs before accepting changes, verify facts and citations after rewriting, and avoid using it to bypass academic, workplace, publishing, or disclosure rules.
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
VirusTotal
64/64 vendors flagged this skill as clean.