Back to skill

Security audit

Cognitive Evolution Stages

Security checks across malware telemetry and agentic risk

Overview

This is a text-only coaching skill for diagnosing learning stages, with no hidden execution, credential use, or privileged access.

This skill appears safe to install as a coaching and reasoning aid. Users should understand that it may guide reflective self-assessment and learning-process diagnosis, but it does not appear to access private data or perform actions on its own.

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.

View on VirusTotal

Static analysis

No suspicious patterns detected.