Free UCP visibility test

Can AI shopping agents find your store?

This free UCP Checker tests how clearly your products, prices, variants, and checkout paths are exposed through public UCP and MCP endpoints. One domain, one live technical test, no account required.

The test follows the same public path an agent would use.

Live diagnostic
UCP / MCP

Public endpoints · no sign-in

No permanent storage

01/.well-known/ucpwaiting
02/api/ucp/mcpwaiting
03tools/listwaiting
04search_catalogwaiting
01

Discover

/.well-known/ucp
02

Connect

MCP tools/list
03

Read

search_catalog
04

Act

variant checkout_url

What the score measures

Visibility is more than publishing an endpoint.

A protocol file can exist while product values are missing. The score totals 100 transparent points. Live values from up to three public products receive full credit; declared schema fields without a live sample receive partial credit.

01

Discovery · 15 points

Can an agent locate a public UCP profile or agents.md without prior configuration?

02

Commerce tools · 20 points

Does the declared service answer and expose recognizable UCP commerce tools? A generic MCP endpoint alone receives only partial credit.

03

Catalog access · 15 points

Does a public UCP catalog request return a readable response?

04

Product data · 40 points

Are description, brand, media, price, availability, product URL, SKU, and GTIN populated across the live sample?

05

Checkout handoff · 10 points

Does each sampled product expose a variant-level, merchant-owned checkout URL?

MCP and UCP

Related protocols, different evidence.

MCP

MCP answers: can an agent call a tool?

The Model Context Protocol standardizes how an AI client discovers and calls tools. A working generic MCP endpoint is a useful signal, but it does not by itself describe a complete commerce journey.

MCP documentation
UCP

UCP answers: can commerce happen through it?

The Universal Commerce Protocol adds commerce-specific discovery, capabilities, catalog structures, and checkout handoffs. The test reports MCP and UCP separately so a generic server is never mistaken for a shopping-ready implementation.

UCP specification

What happens during a test

A bounded public check — not a crawl.

Heartly requests only the storefront and documented public discovery paths, caps response sizes, follows safe HTTPS redirects, and samples at most three catalog products when a public search tool is available. Results are not added to the Heartly Deals Hub; a technical cache may retain a scan for up to one hour.

  • No login or private credentials
  • No product import or publication
  • No inventory mutation
  • One-hour repeat-scan cache

UCP visibility test FAQ

What is a UCP visibility test?

It checks the public signals an AI shopping agent can discover and call, then verifies actual field completeness across up to three sample products. The result covers the UCP profile, MCP tool access, catalog readability, product data, and checkout handoff.

Does a high score guarantee visibility in ChatGPT or other assistants?

No. A strong result means the technical path is available and useful. Every assistant still decides which sources and tools it uses, so the score is not a ranking guarantee or certification.

Are products copied into Heartly?

No. This public test does not import or publish catalog products. When a public catalog is available, up to three examples are read to verify the response and may remain in a technical cache for up to one hour.

Why can MCP work while UCP does not?

MCP is a general tool transport. A store can expose a generic MCP server without publishing a UCP profile, commerce capabilities, a searchable catalog, or variant-level checkout URLs.

Can I test Shopify and WooCommerce stores?

Yes. Enter any public HTTPS storefront. The platform signal is informational; the visibility score is based on the public agent and commerce interfaces the store actually exposes.

Turn a readable catalog into active demand.

Heartly connects storefront signals with campaign automation, a public Deals Hub, Competitor Intelligence, and an AI Assistant built for commerce teams.