ClawHub

Accessibility Toolkit

Security checks across malware telemetry and agentic risk

Overview

This accessibility skill is mostly coherent, but it includes under-scoped smart-home safety patterns such as automatic door unlocking and exposed backup-code messaging.

Review and narrow the automations before installing. Avoid automatic door unlocks unless you add explicit opt-in, multi-factor presence checks, alerts, and fail-secure behavior; use private placeholders or secure storage for access codes; and keep broad voice commands away from locks, alarms, purchases, and emergency actions.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger phrase "Help" is overly broad for a voice-first system and is likely to be spoken in ordinary conversation, media playback, or by bystanders. In this skill context, it is explicitly mapped to immediate attention with no confirmation dialogs, which raises the chance of accidental or unauthorized activation during safety-relevant workflows.

Missing User Warnings

High
Confidence
97% confidence
Finding
The arrival automation automatically unlocks the front door when the person enters the home zone, but the skill does not include a warning about the security implications or guidance on safe deployment. Geofencing errors, device spoofing, shared-account issues, or location drift could cause unintended unlocking, creating a physical security risk to the user and household.

Ssd 3

Medium
Confidence
95% confidence
Finding
The error-reporting example includes a literal manual backup code ("4821") in agent-facing guidance, normalizing the practice of exposing sensitive credentials in messages. In this accessibility context, concise agent messages may be surfaced through notifications, logs, lock screens, speech output, or shared devices, increasing the risk of credential disclosure to unintended parties.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal