maiat-trust-check

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

• Third-party content exposure detected (high risk: 0.80). The skill queries the public Maiat API (e.g., https://app.maiat.io/api/v1/trust?address=) and explicitly uses that third‑party verdict to decide whether to proceed with swaps, so external, public content can directly influence agent actions.

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

• Direct money access detected (high risk: 1.00). The skill explicitly references crypto wallet operations and a paid on-chain API: it requires a funded Base wallet, instructs creating/using a MoonPay CLI wallet (mp wallet create / --wallet <wallet-name>), and shows an mp x402 request that pays $0.02 USDC on-chain. Those are concrete crypto/payment interactions (wallet usage and on-chain payment) rather than generic tooling. It also situates itself as part of a swap workflow and links to a swap skill. Therefore it provides specific crypto/financial execution capability.