ucoz-shop-optimizer-skill

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The workflow reads store products, categories, orders, comments, and customer-facing store content via MCP shop tools, which can expose the agent to user- or customer-submitted strings (such as product names, descriptions, or comments).

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 0.70). The skill operates an e-commerce platform via MCP tools including direct cart management, order submission (checkout_submit), balance operations (balance_change, balance_set), and invoice updates. These are direct financial/commercial operations (purchases, balance adjustments, order placement). While safety instructions and confirmation parameters (confirm: true) are documented in the prompt guidelines, there is no documented service-enforced independent authentication or authorization mechanism (like a validated independent API key or session gating the MCP tool execution layer) for these direct financial actions. Therefore, it falls under direct operations without established qualifying independent control.