Now in Developer Preview NEW

The checkout standard for AI agents.

Trust UCP orchestrates verification, policy, and settlement for agent-driven commerce. One protocol to prove identity, enforce risk, and unlock instant checkouts.

Join the Waitlist See How It Works
JT
MK
AR
500+ developers on waitlist
SOC 2 Type II Ready
Enterprise SLAs
UCP Gateway — Live Transaction
Agent Checkout Request
req_7f3d9a2b1c • Just now
Processing
--ms
Policy Check
Pending
Merchant KYB
99.9%
Gateway Uptime
Discover Reading trust profile...
1
Verify Validating KYB status...
2
Checkout Processing session...
3
Prove Generating audit proof...
4
Waiting for transaction...
99.99%
Uptime SLA
Enterprise-grade reliability
<50ms
Latency P99
Edge-deployed globally
10M+
API Calls/Day
Processed seamlessly
150+
Edge Locations
Worldwide coverage

Powering agent commerce across top marketplaces

amazon
eBay
Walmart
Target
Instacart
DoorDash
Stripe
amazon
eBay
Walmart
Target
Instacart
DoorDash
Stripe
The Coming Chaos

Billions of transactions. Zero verification.

AI agents will transact trillions—but there's no way to know who's behind them, what they're allowed to do, or how to stop them when things go wrong.

Network: UNCONTROLLED
Tap nodes for details
Unverified
No Policy
2.3M/s
Loading network...

Select a node to see why it's a risk

Click any node in the network above

How It Works

Assertions, not PII. Trust, verified.

Trust UCP transmits verified facts about merchants — not sensitive data. AI agents evaluate standardized assertions to make autonomous trust decisions.

Trust Assertion Checker

Evaluating merchant: shop_abc123

Legal Entity legal_entity_exists

Verified legal entity

✓
Registration registration_valid

Current & valid registration

✓
Address address_verified

Physical address verified

✓
Website website_verified

Domain ownership confirmed

✓
Principals principals_verified

Officers/directors verified

✓

Trust Framework Levels

Progressive assurance for different transaction types

basic

Low-value transactions

legal_entity_exists website_verified
standard

Standard e-commerce

+ registration_valid + address_verified
enhanced Active

High-value, B2B

+ principals_verified
premium

Regulated industries

+ ongoing_monitoring
business_verification response 
{
  "status": "verified",
  "trust_framework": "dev.ucp.kyb_v1",
  "assurance_level": "enhanced",
  "assertions": [
    "legal_entity_exists",
    "registration_valid",
    "address_verified",
    "principals_verified"
  ],
  "provider": {
    "id": "com.trustucp"
  }
}
Cryptographic Foundation

Trust you can mathematically verify

No blind trust. No black boxes. Every assertion is cryptographically signed, selectively disclosed, and anchored to public infrastructure.

Privacy-First

SD-JWT Credentials

Selective disclosure lets merchants reveal only the assertions needed—never raw PII. Agents verify claims without seeing underlying data.

Immutable

Hedera Anchoring

Verification proofs are anchored to Hedera's public ledger—immutable timestamps that anyone can audit without trusting a central authority.

Verifiable

Ed25519 Signatures

Every assertion is cryptographically signed. Agents verify authenticity instantly—no API calls, no round trips, no trust assumptions.

How it works together

1

Merchant gets verified

KYB checks generate signed assertions about business identity, registration, and address.

2

Proof anchored on-chain

A hash of the verification bundle is written to Hedera—creating an immutable, timestamped record.

3

Agent verifies locally

AI agents receive SD-JWT credentials, verify signatures, and check anchors—all without exposing PII.

verification_proof.json 
// SD-JWT Verifiable Credential
{
  "type": "MerchantVerification",
  "assertions": {
    "legal_entity_exists": true,
    "registration_valid": true,
    "address_verified": true
  },
  "issuer": "did:hedera:mainnet:...",
  "proof": {
    "type": "Ed25519Signature2020",
    "anchor": "0.0.1234@1706..."
  }
}
Protocol Stack

Everything you need for policy-native commerce

A composable stack built for AI agent commerce. Every layer is programmable, auditable, and ready to scale.

Trust Console

Merchant mission control for verification, policies, and audit proofs. Manage your trust posture from a single dashboard with real-time analytics.

VerificationPoliciesAnalytics

Control Plane

Multi-tenant identity, KYB/KYC, and policy configuration for agent routing at scale.

UCP Gateway

Edge-grade routing for checkout sessions, escalations, and proofs with 99.9% availability.

Adapter Mesh

Connect Shopify, WooCommerce, custom carts, and emerging AI marketplaces seamlessly.

Policy Engine

Define risk tiers, verification rules, and escalation paths for every agent interaction.

For Developers

UCP-native APIs built for AI orchestration

Generate checkout sessions, enforce verification policies, and deliver trust proofs with a single API surface. Built for orchestration engines and agent frameworks.

  • JSON-first, webhook-ready responses
  • Deterministic policy enforcement
  • SDKs for Python, TypeScript, Go
  • OpenAPI spec + Postman collection
Request API Access View on GitHub
checkout.ts 
const session = await ucp.checkout({
  merchant: "shop_abc123",
  agent: "agent_xyz789",
  items: cart.items,
  policy: "standard"
});

// Response includes trust signals
if (session.verified) {
  await session.complete();

  // Access the proof
  const proof = session.trustProof;
  console.log(`Verified: ${proof.merchantId}`);
}
200 OK Response in 47ms
TypeScript
Live API
Why Now

The agentic commerce wave is here

🚀

Agents entering production

OpenAI Operator, Google Agent2Agent, and browser automation tools are shipping now.

🔓

No commerce-specific protocol

MCP and A2A focus on orchestration — none solve merchant verification at checkout.

🏁

First-mover advantage

Define the standard before competitors. Become the default trust layer for AI commerce.

Ready to build trusted agent commerce?

Join the developer preview and help shape the protocol.

We'll reach out with API access and documentation.