ClawHub
Back to skill

Security audit

Agent Tests

Security checks across malware telemetry and agentic risk

Overview

This skill is a local agent-behavior test tracker with disclosed workspace file storage and no evidence of hidden network, credential, or destructive behavior.

Installers should expect this skill to create and update local JSON test-history files in the workspace. Use the --dir option carefully because it directs where those files are read and written.

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.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
tests/run-self-tests.js:157