Back to skill

Security audit

Local Booking

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local-services booking helper that can share contact details with Lokuli only as part of user-confirmed booking flows.

Before installing, understand that this skill connects to Lokuli to search providers and create bookings, and booking requires sharing contact and location details. Confirm provider, time, price, and what information will be sent before approving any booking or payment link.

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
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger phrase 'any local service request' is overly broad for a skill that can initiate real-world bookings and collect user contact information. Broad activation increases the chance the skill is invoked in ambiguous contexts, causing unnecessary disclosure of location/contact data or steering users into transactional flows they did not explicitly request.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill instructs collection of name, email, phone, and ZIP code and transmits them to a third-party MCP endpoint, but it does not provide a clear privacy notice or consent step before doing so. In a booking context involving real-world services, this omission raises meaningful privacy and compliance risks because users may not realize their personal data is being shared externally.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
## Rules

- **Never book without confirmation** — Always get explicit approval
- **Show pricing upfront** — Use get_pricing_estimates if needed
- **Collect required info** — Name, email, phone before booking
- **Default to user's ZIP** — If known from context
Confidence
75% confidence
Finding
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Static analysis

No suspicious patterns detected.