Oris runs automated safety checks on every transaction before execution. The screening evaluates counterparty risk, sanctions exposure, and transaction patterns. If a check fails, the transaction is blocked. Your application receives a clear error code. No unsafe payment reaches the provider.
Your agent calls agent.pay(). Oris receives the transaction request with the destination address, amount, and chain identifier.
agent.pay(
to="0xA1b2...C3d4",
amount=25.00,
chain="base"
)The safety engine evaluates the counterparty address, the transaction amount, and the behavioral context against the agent's history. This evaluation completes in under 100 milliseconds.
If the check passes, the transaction proceeds to your payment provider. If the check fails, Oris returns a structured error with the failure reason. The transaction never reaches the provider.
The safety engine operates on a fail-closed principle. If the screening service is unavailable, transactions are blocked. If the evaluation times out, transactions are blocked. If any component in the safety pipeline returns an ambiguous result, the transaction is blocked.
This architecture guarantees that system failures produce safety rather than exposure. A degraded system will refuse transactions until full service is restored.
The safety screening engine is powered by Veris, the stablecoin compliance infrastructure built by the same team. Veris indexes 10 blockchains in real time, maintains 70,000+ labeled addresses, and runs 15 ML models for pattern detection.
This infrastructure runs silently behind every Oris transaction. Your agents benefit from enterprise-grade compliance without any configuration.
The destination address is checked against known risk indicators, sanctions lists, and behavioral patterns. Addresses associated with illicit activity are flagged before the transaction reaches your provider.
Unusual amounts, frequencies, or destinations trigger additional review based on the agent's behavioral baseline. The system compares each transaction against the agent's established norms.
The screening engine considers the agent's activity across all 10 supported blockchains, providing a complete risk picture. A single-chain view is insufficient for modern threat detection.
Automated safety checks are included in every Oris plan. No configuration required. Every agent payment passes through the screening pipeline from the first API call.