Six checks, each on a public file or page that AI crawlers and shopping agents read. The checker fetches them the way a crawler does, from the outside, and makes no calls to any AI model.
AI crawlers in robots.txt
What we look for
Your /robots.txt rules for 9 AI crawlers: OAI-SearchBot, ChatGPT-User and GPTBot (OpenAI), Claude-User and ClaudeBot (Anthropic), PerplexityBot, Google-Extended, Applebot-Extended and CCBot (Common Crawl). For each one we apply the group that names it, or else the * group, to your home page and a product page, as RFC 9309 says crawlers must.
Why it matters
Search and user crawlers fetch pages to answer questions and cite sources right now: block them and you can’t be quoted or linked. Training crawlers gather data for future models, which is a business choice. The checker flags the first as a problem and the second as a warning.
How to fix it
On Shopify, edit robots.txt with a templates/robots.txt.liquid file and remove the Disallow rules for the crawlers you want. If your domain runs through Cloudflare, also check its “Block AI bots” setting, which turns crawlers away before they read robots.txt.
Sources
llms.txt
What we look for
A text file at /llms.txt in the llmstxt.org format: an H1 title (the one required part), an optional “>” summary, and “##” sections with Markdown links. We open up to 6 of its links to check that they work.
Why it matters
llms.txt gives assistants and agents a short, curated map of your store instead of making them crawl it. On Shopify the same file carries the links to the store’s agent endpoints.
How to fix it
On Shopify, add a templates/agents.md.liquid file (Shopify serves it at /agents.md and /llms.txt) with your store description, key collections, policies and FAQ, and keep the UCP and MCP lines from the agents object. Elsewhere, upload a plain llms.txt at the root of your site.
Sources
Structured data
What we look for
JSON-LD on the home page (Organization and WebSite) and on one product page (Product with offers, price, availability and brand, and a GTIN if you have barcodes). Blocks that aren’t valid JSON are flagged.
Why it matters
Structured data states facts in a form machines don’t have to guess: who sells, what it costs, whether it’s in stock. Search engines use it for rich results, and AI systems use it to read product facts reliably.
How to fix it
Most Shopify themes output product data with the structured_data filter: check that yours does, and add Organization and WebSite on the home page. Fill in the vendor and barcode fields on your products.
Sources
Key text in the HTML
What we look for
Whether the product’s title, price and description are visible text in the first HTML response, before any script runs. Without a product page, we check that the home page has real text and an h1 heading.
Why it matters
Most AI crawlers read the raw HTML and don’t run JavaScript. Text that only appears after a script runs (some page builders, review and price apps, headless storefronts) is invisible to them.
How to fix it
Render key content on the server. In a Liquid theme, output the product title, price and description in the section’s HTML, not through an app script. For a headless storefront, use server-side rendering.
Sources
Sitemap
What we look for
A valid XML sitemap at /sitemap.xml, and a “Sitemap:” line in robots.txt that names it.
Why it matters
The sitemap lists every page you want found, with update dates. Crawlers use it to discover products and to know when they changed.
How to fix it
Shopify builds /sitemap.xml and names it in robots.txt on its own: if yours fails, check that the store isn’t password-protected. On other platforms, turn on the sitemap in your SEO settings and add the “Sitemap:” line.
Sources
Shopify agent endpoints (UCP)
What we look for
Whether the site runs on Shopify, and whether it publishes a Universal Commerce Protocol (UCP) profile at /.well-known/ucp with an MCP endpoint and the catalog, cart and checkout capabilities. On Shopify we also sample /products.json for prices, images, descriptions and brands.
Why it matters
AI shopping agents use these endpoints to search a store, build a cart and send the shopper to checkout without scraping pages. On Shopify they come with the platform; complete product data makes your products easier to match and recommend.
How to fix it
On Shopify, take the store off its password page, keep titles, descriptions, images, prices, vendor and barcodes complete, and keep the UCP lines in robots.txt and agents.md. On another platform, ask whether it supports UCP.
Sources
