ClawHub
Back to skill

Security audit

Agent Safety Monitor

Security checks across malware telemetry and agentic risk

Overview

The skill does not show theft or destructive behavior, but it overstates real-time safety enforcement and references missing components, so it should be reviewed before use.

Treat this version as a prototype offline log analyzer, not a complete guardrail system. Do not rely on it to actually pause or kill agents, enforce file/API/network access, run a dashboard, or handle wallets or sensitive credentials unless those capabilities are added in reviewed artifacts.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Credential Access

High
Category
Privilege Escalation
Content
# Blocked target patterns
blocked_targets:
  - /etc/shadow
  - .env
  - private_key
  - password
Confidence
95% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Credential Access

High
Category
Privilege Escalation
Content
# Blocked target patterns
blocked_targets:
  - /etc/shadow
  - .env
  - private_key
  - password
Confidence
60% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.