ClawHub
Back to skill

Security audit

sold-price-comps

Security checks for vulnerabilities and agentic risk

Overview

This paid MCP skill is not clearly malicious, but it needs review because it can automatically charge per call and its sold-prices workflow currently shows active asking-price data.

Before installing, confirm your MCP client requires approval or has a spending cap for x402 payments, and treat results marked active as asking prices rather than completed sales. Avoid sending confidential inventory, customer, or business strategy details in product queries because calls are logged by the provider.

SkillSpector

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

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The example response contradicts the skill’s stated purpose of providing recent sold prices by returning `data_type: active`, `source: eBay (active)`, and null `sold_date` values. This can mislead downstream agents or users into treating asking prices as completed-sale comps, causing inaccurate valuation, pricing, or purchasing decisions in a pay-per-call marketplace context.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill explicitly states that x402-capable clients handle payment automatically and that calls are pay-per-call, but it does not present a prominent user-facing warning that invoking the tool can trigger real monetary charges. In an agent context, this can lead to unintentional purchases or repeated paid requests without meaningful user consent, especially when agents autonomously select tools.

External Transmission

Medium
Category
Data Exfiltration
Content
{
  "mcpServers": {
    "sold-price-comps": {
      "url": "https://api.3rdplaceprovisions.com/mcp"
    }
  }
}
Confidence
80% confidence
Finding
The skill directs the agent to transmit user queries to an external third-party MCP endpoint, and the file also notes that every call is logged server-side. While external connectivity is necessary for the skill's function, this creates a real data-exposure risk because user-provided product descriptions, usage metadata, and potentially sensitive inventory/business information are sent to and retained by a remote provider.

Static analysis

No suspicious patterns detected.