PLENA AI Agent Action Registry

What it records

A registry for agentic work: not just what the AI said, but what it was authorized to do and who remains accountable.

Agent identity

Agent name, system owner, business unit, version, tool scope, and action boundary.

Authority basis

Policy, contract, delegated role, ticket, workflow, customer consent, or board-approved basis.

Human review

Reviewer, approval, rejection, escalation, reversal, refusal, and correction path.

Evidence trail

Source systems, logs, evidence hashes, retention period, redaction level, and verification URL.

Question Fortune 500 buyers ask	PLENA answer
Which AI agents exist and which ones acted?	Agent Action Registry entry with actor type, name, owner, and task scope.
Could the agent approve, spend, deny, terminate, publish, or transmit?	Authority basis plus explicit scope limits.
Was a human responsible?	Human review object and reviewer identity reference.
Can legal, compliance, or audit reconstruct what happened?	VRX-1 Enterprise receipt with evidence references and verification URL.

Sample Agent Action Receipt JSON

{
  "protocol": "VRX-1",
  "profile": "enterprise-evidence",
  "receipt_id": "PLENA-EAR-20260514-0001",
  "issuer": {
    "name": "Example Enterprise",
    "organization_id": "ORG-EXAMPLE-001",
    "issuer_role": "AI Governance Office"
  },
  "issuing_product": "PLENA AI Agent Action Registry",
  "receipt_type": "ai_agent_action",
  "created_at": "2026-05-14T17:00:00Z",
  "subject": "AI agent action authorization record",
  "summary": "Example record showing that an AI agent action was authorized for a bounded enterprise task and routed for human review.",
  "status": "draft",
  "privacy_level": "private",
  "sharing_mode": "audit_room",
  "jurisdiction": "example",
  "actor": {
    "actor_type": "ai_agent",
    "actor_name": "Procurement Review Agent",
    "actor_id": "AGENT-EXAMPLE-007",
    "system_owner": "Procurement Operations"
  },
  "authority_basis": {
    "basis_type": "policy",
    "basis_reference": "AI-USE-POLICY-EXAMPLE-2026",
    "scope_limit": "May summarize vendor questionnaires; may not approve vendors or alter contract terms."
  },
  "human_review": {
    "required": true,
    "status": "pending",
    "reviewer_id": "REVIEWER-EXAMPLE-112"
  },
  "risk": {
    "risk_level": "high",
    "risk_domains": [
      "vendor",
      "contract",
      "customer_data"
    ],
    "control_mapping": [
      "NIST AI RMF: Govern",
      "ISO/IEC 42001: AI management system evidence",
      "Internal AI use policy"
    ]
  },
  "evidence": {
    "evidence_hashes": [],
    "source_systems": [
      "ServiceNow",
      "Microsoft Purview",
      "Jira"
    ],
    "retention_period": "7 years or legal hold",
    "redaction_level": "private"
  },
  "verification_url": "https://joinplena.com/verify.html?receipt=PLENA-EAR-20260514-0001",
  "related_receipts": [],
  "appeal_or_correction_path": "Internal AI governance escalation queue"
}