Agent-to-Agent (A2A) Communication
Carry verifiable interaction records through A2A metadata. Prove authorization across agent hops.
Overview
Agent-to-Agent (A2A) communication enables autonomous agents to interact directly with each other while maintaining verifiable transaction history. Originary provides PEAC-Receipt attach points that allow agents to prove prior interactions and policy compliance.
When Agent A calls Agent B, Originary issues a signed receipt. Agent B can then attach this receipt when calling Agent C, creating a verifiable chain of agent interactions that can be audited and validated.
Use Cases
Multi-Agent Workflows
Coordinate complex workflows where multiple agents collaborate, each maintaining receipts of their contributions.
Policy Enforcement
Verify that agents have obtained necessary permissions or payments before proceeding with downstream operations.
Audit Trails
Maintain complete, verifiable history of agent interactions for compliance and debugging purposes.
Implementation
Attaching PEAC-Receipts
# Agent B calling Agent C with attached receipt curl https://agent-c.example.com/api/action \ -H "Authorization: Bearer agent-b-token" \ -H "PEAC-Receipt: eyJhbGc...agent-a-to-b-receipt"
Agent C can verify the attached receipt to confirm that Agent B received proper authorization from Agent A before executing the requested action.
Benefits
- Each agent-to-agent call produces a signed receipt proving what was requested and delivered
- Receipts chain together: Agent C can verify the full path back through B to A
- Offline verification using Ed25519 public keys; no callback to issuer required
- Receipt fields include resource, timestamp, policy hash, and optional payment evidence
- Standards-based approach compatible with existing protocols