ClawHub
Back to skill

Security audit

book anything

Security checks across malware telemetry and agentic risk

Overview

This skill is mostly transparent, but it can place paid AI phone calls and change real reservations without a clear required confirmation step.

Install only if you trust LifeClaw with booking details and paid outbound calls. Configure your agent to ask for explicit confirmation before every phone call, cancellation, reschedule, booking-record update, or disclosure of contact details, and monitor/revoke the LIFECLAW_API_TOKEN through the Telegram bot if needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Vague Triggers

High
Confidence
94% confidence
Finding
The skill advertises very broad trigger phrases such as "find," "book," "reserve," "cancel," and "reschedule," which are common words in ordinary conversation. That creates a real risk of unintended invocation for a high-impact skill that can search businesses and initiate phone-based actions on a user's behalf, potentially causing data transmission or external side effects without sufficiently specific user intent.

Natural-Language Policy Violations

Medium
Confidence
84% confidence
Finding
The description states that the AI caller automatically matches the merchant's language, and the body later says language is auto-inferred and should not be set. This removes user control over how the call is conducted and may lead to calls being placed in a language the user did not expect, increasing the chance of miscommunication, privacy surprises, or unauthorized-seeming actions.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The API exposes real-world phone-call actions that can place bookings, cancellations, reschedules, and inquiries with per-minute charges, but the referenced section does not provide a prominent safety warning or confirmation requirement at the action boundary. In an agent setting, this increases the risk of unintended real-world side effects and surprise financial cost if a model or integrator triggers `/skill/book/phone` from ambiguous user input.

VirusTotal

46/46 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.