Prebuilt risk-score thresholds for every action type in the Core Ontology.
Scan the product manifest — ontology types, functions, actions, policies, and dataset shells.