ClawHub
Back to skill

Security audit

Homeassistant Mcp

Security checks across malware telemetry and agentic risk

Overview

This Home Assistant skill is mostly purpose-aligned, but it needs review because it directs agents to use raw shell commands with a long-lived token for sensitive smart-home camera and state access.

Install only if the Home Assistant endpoint is yours and you are comfortable giving the agent a long-lived token. Use the least-privileged token possible, prefer MCP tools over shell REST fallbacks, require explicit confirmation before camera or garage actions, and delete saved snapshots when no longer needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs the agent to use generic shell execution with curl to access authenticated Home Assistant camera endpoints and save files locally. This expands the skill from constrained MCP-mediated operations into arbitrary command execution and direct network access, which materially increases the chance of command abuse, unintended file writes, token leakage, and privacy-sensitive camera access.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The camera snapshot workflow directs authenticated retrieval of potentially sensitive in-home images without requiring an explicit user warning or consent flow. In the context of a smart-home skill, camera feeds are highly privacy-sensitive, so normalizing silent token-authenticated snapshot access materially raises surveillance and data exposure risk.

Missing User Warnings

Low
Confidence
84% confidence
Finding
The documented REST fallback performs authenticated requests to the Home Assistant API without any user-facing disclosure that data is being transmitted to and retrieved from a network service. While this is less sensitive than camera snapshots, it still exposes internal home-state data and normalizes network operations outside the declared MCP interface.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.