ClawHub
Back to skill

Security audit

Ao Op

Security checks across malware telemetry and agentic risk

Overview

This skill is a transparent local wrapper for an Agent Orchestrator checkout, with disclosed commands that can change local files or running sessions.

Install this only if you intend to operate the AO checkout at the documented local path. Treat start, stop, send, update, git pull, pnpm install, and pnpm build commands as deliberate actions because they can affect running sessions, local source files, dependencies, and generated build output.

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
Findings (1)

Missing User Warnings

Low
Confidence
93% confidence
Finding
The skill documents `git pull`, `pnpm install`, and `pnpm -r build` as a maintenance flow without an explicit warning that these commands modify the local checkout and may execute install/build scripts from the repository. In a skill intended to be invoked by an agent, that omission can lead to unintended state changes or supply-chain exposure if run without clear user confirmation.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.