ClawHub
Back to skill

Security audit

Book Fence

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent fence-booking integration, but it can send a customer's name, email, and phone number to an external booking service without explicit consent or confirmation instructions.

Before installing, treat this as a transactional booking skill: only use it when you intend to contact or book a fence service, and require the agent to show the provider, time slot, and exact contact details before submitting anything to Lokuli.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

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 (2)

Missing User Warnings

High
Confidence
98% confidence
Finding
The skill includes a booking flow that collects and transmits personal contact information to a third-party MCP endpoint without any user-facing warning, consent language, or data-handling notice. Because the data includes name, email, and phone number and is sent to an external service, users may unknowingly disclose sensitive personal information and the agent could violate privacy expectations or organizational policy.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger language is broad enough that the skill could activate on general discussion about fences rather than a clear booking intent. In a transactional skill that can search providers and create bookings, accidental invocation increases the chance of unintended data collection, external API use, or steering the conversation into a purchasing flow the user did not explicitly request.

Static analysis

No suspicious patterns detected.