agentpay

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly returns API keys and one-time card PAN/CVV values (shown once with "SAVE THIS") and the agent would need to display or relay those secrets verbatim to the user, creating an exfiltration risk.

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

• Direct money access detected (high risk: 1.00). The skill is explicitly a payment service: it issues API keys, accepts funding via a checkout link, reports balances, and — critically — creates single-use virtual payment cards (PAN/CVV/exp) via POST /spend that can be used to execute payments. This is a purpose-built financial API (issuing spendable card details and managing funds), not a generic tool. Therefore it grants direct financial execution capability.