ClawHub
Back to skill

Security audit

OpenClaw Security Guard

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed local security-check skill; its scripts inspect user-selected content and optionally write local audit/helper files, with no evidence of hidden network exfiltration or destructive behavior.

Use this as a lightweight preflight scanner, not proof that code is safe. Run audits only on folders you intend to inspect, treat generated JSON and Obsidian notes as sensitive because they may include excerpts from scanned files, and run the hook installer only if you want a persistent local prepublish helper.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill explicitly instructs users to run local shell commands (`node ...`, `bash ...`) but the metadata declares no permissions. This creates a trust and review gap: consumers or platforms may underestimate the skill's execution capabilities and allow shell-backed actions without appropriate scrutiny or sandboxing.

VirusTotal

66/66 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
scripts/prepublish-guard.mjs:13