text-to-avatar

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

• Third-party content exposure detected (low risk: 0.10). SKILL.md describes a text_to_avatar tool that takes user-supplied prompt and optional reference images URLs as inputs, but it does not specify that the runtime fetches or reads arbitrary outsider-authored free text from any third-party feed/queue/inbox without first selecting a specific user-provided item.

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

• Direct money access detected (high risk: 1.00). The skill exposes a paid tool ("text_to_avatar") that requires a funded wallet (MCP_SVM_PAYMENT_ADDRESS) and states calls are "settled automatically" in USDC (x402). It specifies a per-call price ($0.15 USDC) and will return a PaymentRequired structured result when no payment payload is attached. These are explicit payment/crypto settlement mechanics (wallet address, automatic settlement, and payment payloads), which constitute direct financial execution capability.