API

What consequence boundary is this action about to cross?

The evaluate endpoint classifies consequence topology and returns the governance intensity required before execution.

POST/api/evaluate

Example request

{
  "packet_id": "wire-transfer-t0",
  "system_context": {
    "domain": "finance",
    "workflow": "payment-release",
    "autonomy_level": "agentic"
  },
  "requested_action": {
    "type": "release_payment",
    "execution_surface": "wire_transfer",
    "reversibility": "irreversible"
  },
  "consequence_profile": {
    "level": "critical",
    "propagation_targets": ["bank", "vendor", "ledger"],
    "binding_moments": ["settlement"],
    "blast_radius": "multi_party"
  },
  "authority_state": { "status": "valid" },
  "reality_contact": { "status": "confirmed" },
  "safeguards": { "operator_review_confirmed": true }
}

Outcomes

  • BOUNDARY_LOW — standard governance intensity.
  • BOUNDARY_ELEVATED — revalidation/escalation recommended.
  • BOUNDARY_CRITICAL — execution should be blocked until missing safeguards are satisfied.