Registry

Behavioral standard for clinical-decision-support and care-navigation agents operating in PHI-adjacent contexts. Input hashes are sealed; decision metadata is sealed.

Behavioral standard for financial-services agents handling transactions, risk scoring, and compliance workflows. Severity required; decision metadata sealed.

Behavioral standard for legal AI agents performing document review, contract analysis, and legal research. Hashes are recorded; decision metadata is sealed.

Default behavioral standard for general-purpose AI agents. Records action type, tool invocation, and hashed payloads with notice-level visibility.