How It Works
Prime Form Calculus sits between AI systems and real-world execution. It evaluates policy, verifies authorization at decision time, and creates deterministic audit records for every governed action.
Organizations can request API access through the developer signup flow and start integration from there.
PFC receives the requested action, the target resource, and the execution context before anything irreversible happens. The critical control point is the execution boundary, where actions move from proposal to commitment.
This is also where execution risk in AI systems becomes real, because a decision only becomes consequential when a downstream system is about to act on it.
PFC checks whether the action is allowed under the active policy and whether the caller is still authorized to proceed. This is the enforcement layer behind execution control at the execution boundary.
See how PFC enforces this at the execution boundary → Security
The API returns an allow or block decision with a reason code and a stable decision identifier for downstream handling.
Each governed decision produces a deterministic evidence record so teams can audit what happened without relying on ad hoc logs.
Teams that want the full explored-to-executed chain can see a verifiable AI decision example in the public Creative Lineage Demo.
See the Security page for the execution-boundary framing behind this control point.
Every governed action produces a signed receipt. Downstream systems can verify that receipt locally and reject actions without valid proof.