What your agent endpoint serves
This is the only check here that reads what an AI agent actually receives rather than what your pages contain. We ask your store for its catalogue the way a shopping agent would, and report what came back.
Across 10,110 Shopify storefronts we have scanned, 25% fall short on the catalogue their agent endpoint serves.
Or read it off your own store. The scan fetches your storefront the way a crawler does and answers this question and seven others about it.
Free, no signup. Takes about ten seconds.
What we did
Your store answers an agent-commerce endpoint at
/api/ucp/mcp, speaking the Universal Commerce
Protocol. You did not switch it on and it is not in your admin. We called it, asked
for products matching one of your own product titles, and read the fields it returned.
You can do the same in one command —
we show you how.
Most of what comes back is Shopify’s doing and is already correct: the title, the description, the canonical URL, the price as an integer with a currency, the variants, the images. There is nothing to fix there and we will not invent something.
The field that is yours
Product category. In your Shopify admin, on the product, the Category field — it maps to Shopify’s standard product taxonomy and it travels to the agent as a category identifier. A product with none arrives at an agent uncategorised.
What that costs you is worth being exact about, because the obvious answer turns
out to be wrong. The protocol has a category filter —
filters.categories is in the published schema, so an agent is plainly meant
to be able to narrow a search to a category. We tested that on
200 live Shopify stores on 2 September 2026.
Of the 190 that answered, none of them
filtered: 186 returned the full unfiltered result set
for a category that does not exist, and 4 returned
nothing for any category at all, including the one their own products carry. In the same
request shape a price.max of one cent worked on
150 of them, so this is the field, not our call.
Whatever that filter is doing today, it is not narrowing a search to your category, and
we are not going to tell you a blank one gets you excluded from a set.
Every reading is here, store by store.
What it does cost you is smaller and real. The category is the only field in the whole response that says what the product is in a form a machine can act on without interpreting prose. Your title and description are sentences; an agent deciding whether your item answers the question in front of it has to read them and infer. The category is the one line it does not have to guess at, and on a fifth of products it is blank.
It is set per product, it takes seconds, and Shopify will suggest the category from the product title if you let it. If you have hundreds of products, the bulk editor takes the column.
SKUs, which we report and do not mark you down for
If a store returns no SKU on any variant we say so, because an agent reconciling an order it placed last month, or matching your item against a price comparison, uses it. But plenty of stores have no business having SKUs — one-off vintage stock, made-to-order work, a single-product store — and penalising those would be scoring a store down for its business model. So it is a note, not a finding.
What this check is not
It is not a measure of whether AI can find your store. The endpoint answers an agent that already knows your domain and is asking you directly — the buying path, which Shopify has covered.
Getting named in an assistant’s answer is a different problem that runs through your pages rather than this endpoint, and it is decided by what your product pages state literally and by what other people have written about you. Two paths. This check measures one of them and says which.
If this check is missing from your report, it means we could not get a clean reading — a rate limit, a timeout, or a query of ours that matched nothing in your catalogue. That is our problem rather than yours, and we would rather show nothing than a guess.
Checking your own store
The scan below reads your storefront the way a crawler does and reports this check along with the rest. It is free and needs no account.
Free, no signup. Takes about ten seconds.
The other checks
- Product structured data on Shopify
- Allowing AI shopping crawlers on Shopify
- When your product page needs JavaScript to read
- llms.txt on Shopify
- When your Shopify sitemap is missing
- Page titles and descriptions on Shopify
- When your homepage is a JavaScript redirect
Installs on any Shopify store. The free plan needs no card, and the dashboard runs on sample data if you would rather look first.