Security audit

Self-Criticism

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed self-review workflow that keeps small local memory files and does not show hidden network, credential, destructive, or deceptive behavior.

Install only if you want persistent self-critique behavior. Review any proposed writes to main memory, ~/self-criticism/, AGENTS.md, or SOUL.md, and keep saved notes limited to process lessons rather than sensitive personal or project details.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
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)

Missing User Warnings

Low

Confidence: 91% confidence
Finding: The template includes shell commands that create directories and files in the user's home directory without any explicit warning, consent step, or safer alternative. Even though the commands are simple and not overtly destructive, agent skills can cause unintended filesystem modifications if executed automatically, which violates least surprise and can be risky in security-sensitive environments.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill instructs the agent to persist behavioral preferences to main memory and local files across sessions without clearly requiring a user-facing consent step or warning about retention. This can create privacy and autonomy risks by storing user behavioral patterns durably, potentially surprising users and influencing future sessions in ways they did not knowingly approve.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.