ClawHub
Back to skill

Security audit

IoT Platform Connection

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only skill, but it shows users how to put a powerful OpenClaw MCP control endpoint on the public internet with insufficient safety guidance.

Install only if you understand OpenClaw gateway security. Prefer loopback or Tailscale Serve/tailnet-only access. Do not use Funnel for this gateway unless you add strong access controls, restrict exposed tools, use a strong unique secret, monitor access, and are comfortable exposing device, channel, and agent-control functions remotely.

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 (1)

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill documents a Tailscale Funnel mode that makes the MCP gateway publicly reachable while the same guide advertises device control, sensor access, channel messaging, and agent execution capabilities. Presenting this configuration without a strong warning, threat model, or hardening guidance can lead users to expose a powerful control plane to the internet with only password-based protection, increasing the risk of unauthorized access and remote abuse.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.