ClawHub
Back to skill

Security audit

Dana

Security checks across malware telemetry and agentic risk

Overview

The skill matches a clinic booking purpose, but it sends phone and appointment details to an external booking API without an in-flow confirmation step and includes hardcoded API tokens.

Install only if you are comfortable with a third-party BeautsGO booking service receiving appointment timing, party size, and any phone number you provide. Before using it for real bookings, prefer a version that asks for explicit confirmation before submission and moves API tokens out of the distributed skill package.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Tp4

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding
The skill metadata and description present it as an informational clinic/booking helper, but the documented behavior includes broader capabilities: multi-turn intent recognition, collection of personal data such as phone numbers and appointment details, external API submission, and opening external pages in the system browser. This mismatch reduces informed user consent and reviewer visibility, which is dangerous because users may disclose sensitive information or trigger external actions they did not expect from the declared scope.

Vague Triggers

Medium
Confidence
74% confidence
Finding
The skill description and intent flow do not define specific activation phrases, scope limits, or guardrails for when booking, consultation, price lookup, or browser-opening behaviors should occur. Ambiguous invocation increases the chance of unintended triggering and accidental collection or transmission of user data, particularly in a skill that performs external actions and handles appointment information.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
The skill collects contact information and appointment details, then transmits them to an external booking API without an explicit user-facing notice at the point of submission or clear consent flow. This creates a privacy/security issue because users may provide phone numbers and scheduling data without understanding that the data leaves the local skill context and is shared with a third-party service.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
api/skill.js:25