Google launched Universal Cart. We have 30 days.

On May 19, Google I/O 2026 announced Universal Cart - an intelligent cart that follows users across Search, Gemini, YouTube, and Gmail. The early checkout participants, per ALM Corp's May 20 analysis: Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, and Shopify brands.

I read that list twice looking for a Magento store. None.

That's the article.

If your store runs on Magento or Adobe Commerce, agent-routed traffic is going to flow past you - first in the US, then Canada and Australia "in the coming months," then the UK. The agent layer isn't going to wait for Adobe Commerce to ship native UCP support. The merchants in the first cohort had thirty days of head-start. Most of that window is already gone.

Here's what to ship before the rest of it closes.

The five-protocol stack, compressed

Four protocols define how an AI agent buys something on behalf of a user. A fifth ties payments together.

• UCP - the discovery layer. Your store publishes a manifest at /.well-known/ucp declaring its capabilities, transports, and payment handlers.
• MCP - the transport layer. Agents dispatch your commerce tool calls over MCP messages.
• ACP - OpenAI and Stripe's checkout protocol. Stripe-led coalition.
• AP2 - Google's payment-authorization protocol. Sixty-plus partners signed at launch: Adyen, American Express, Mastercard, PayPal, Coinbase, Revolut, Worldpay, and more.
• MPP - Stripe's machine-payments protocol. Same family as ACP.

Benji Fisher's synthesis post on dev.to is the sharpest framing I've read: UCP discovers, MCP transports, ACP and AP2 authorize. Read it if you haven't.

The UCP spec itself is densifying fast. A loyalty extension landed on May 19 (#340). A schema-validated documentation harness landed on May 21 (#359), with AP2 mandate examples now machine-validated against the spec. The cadence suggests v1.0 GA in weeks, not months.

You don't have to learn five protocols. You have to make one decision.

The one decision

Pick an ecosystem to plug into first based on the payment processor your store already runs.

The protocol war is coalition-driven, not technically segmented. Pick the side your existing rails are on. You can ship the other later.

The 30-day playbook

Five concrete steps. Each is a Magento-specific task, not a generic "audit your Merchant Center" instruction.

Day 1-3: publish /.well-known/ucp

Drop a JSON manifest at your store root. Skeleton for Magento 2.4.9:

{
  "ucp": {
    "spec": "https://ucp.dev/spec/2026-04-08",
    "schema": "https://ucp.dev/schemas",
    "endpoint": "https://your-store.com/rest/V1",
    "capabilities": ["catalog", "cart", "checkout", "order"],
    "transports": ["rest", "mcp"],
    "payment_handlers": ["ap2", "stripe"]
  }
}

Don't hand-write this and hope. Validate it against the upstream ucp-schema tool (step 5).

Day 3-10: map your REST API to MCP transport

Your Magento /rest/V1 endpoints already implement most of the UCP capability surface - catalog, cart, checkout, order. You're not rebuilding. You're wrapping those endpoints so an agent can call them over MCP's tools/call envelope.

This is where most of the engineering time goes. Plan for a focused two-week sprint with one senior backend engineer who knows your REST customizations. If you've stayed close to the default REST contract, the wrapping is mechanical. If you have heavy REST customizations, budget more.

Day 10-15: declare a payment handler that matches your processor

If you're on Adyen, publish an ap2 payment-handler namespace in your manifest. If you're on Stripe, publish an acp namespace. The handler tells the agent which authorization protocol your checkout speaks.

This is declaration only at the manifest level. The actual payment integration on the Magento side already exists in your payment module - you're advertising it, not rewriting it. The work is in your manifest, not in your checkout code.

Day 15-22: wire loyalty - optional but high-leverage

UCP's loyalty extension shipped this week. If you have a Magento Reward Points module or a custom loyalty integration, declare it in the loyalty extension namespace. Loyalty surfaces in agent UI as a price-impact signal - it's how an agent decides which store to recommend when two have the same product at the same list price.

Stores skipping this in 2026 will lose the price-tied-discount segment for the next three years.

Day 22-30: validate before you ship

A broken manifest costs you visibility. Agents that hit a 404 or a JSON parse error mark you unreachable and skip you on retry - you don't get a second chance until the next crawl cycle.

We built a free AI-readiness audit at webmaster-ramos.com/audit that probes /.well-known/ucp, validates the manifest against the published UCP profile schema, checks transport declarations, and flags missing payment-handler namespaces. It's KB-driven and rule-based - no LLM in the pipeline, no judgment calls, just deterministic checks against the spec.

Run it twice: once before you go live, once after. If you don't fire any of the UCP-readiness rules, you're shippable.

If you'd rather wire your own check, the manifest is a plain JSON document validatable against the UCP profile schema with any standard JSON Schema validator. The schema is published at https://ucp.dev/schemas/profile.json.

What Adobe is doing (and not doing)

Adobe shipped AI agents into AEM eight weeks ago - Experience Modernization Agent, Development Agent, Content Optimization Agent. They're shipping more every release cycle.

Adobe has shipped nothing agentic into Adobe Commerce.

That asymmetry is eight weeks deep and widening. If you're waiting for Adobe to ship native UCP support for Commerce, you're waiting on a roadmap that hasn't surfaced publicly. The merchants who show up in next year's Universal Cart case studies are the ones who built their own UCP layer this quarter.

You have thirty days.