safe

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

• Third-party content exposure detected (medium risk: 0.30). The workflow consumes multisig transaction feeds and details from the Safe Transaction Service API, which can contain outsider-submitted transaction metadata and free text without prior agent selection.

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

• Direct money access detected (medium risk: 0.30). The skill documents programmatic multi-signature transaction execution, transfers, batch calls, and contract deployments via the Safe{Core} SDK (protocol-kit and api-kit). This constitutes direct financial operations (transfers, purchases, trades/swaps or settlement authorization). These operations are protected by independent access controls enforced by the underlying Safe smart contracts and execution layer (multisignature threshold requirements, owner signatures, and cryptographic validation), which require threshold signatures before execution succeeds.