Ontario Protocol.
Agent commerce architecture

Autonomous Agent Commerce

Autonomous agent commerce needs more than a payment rail. Agents need discovery, preflight verification, wallet policy, paid-tool metadata, settlement proof, and a durable way to reject broken endpoints.

autonomous agent commerce Agent platform teams, marketplace builders, and API teams adding paid machine customers.
Explore verified services Use can-pay policy

Commerce loop

The commerce loop is discover, verify, decide, pay, settle, and monitor. Ontario exposes each step as HTML for humans and JSON for agents.

  • Discovery catalog lists paid endpoints and service metadata.
  • Verification reports explain whether an endpoint is ready to receive agent payments.
  • Settlement and drift monitors keep trust from decaying after launch.

Why trust signals matter

Autonomous buyers cannot rely on brand claims alone. They need explicit payment metadata, public reports, structured service profiles, and evidence that can be rechecked before every spend.

  • Machine-readable metadata reduces integration ambiguity.
  • Signed reports make readiness evidence linkable and auditable.
  • Alert routes and weekly intelligence convert failures into follow-up work.

Workflow

Publish a paid endpoint with explicit x402 payment requirements.
Expose service metadata through manifest, OpenAPI, and discovery profiles.
Run readiness verification and save a public report.
Wire agent policy to deny stale, unsigned, or price-mismatched endpoints.
Monitor reports, settlements, and ecosystem findings over time.

Use these Ontario routes to move from content research into live verification evidence.

Discovery catalog Agent payment preflight Weekly intelligence