Shopify and AI search: what crawlers meet, and what to change
What AI crawlers meet on Shopify
Shopify manages robots.txt: one group for every crawler, allowing the store and keeping carts, checkout, accounts, sorted collections and URLs with two or more filters out. No AI crawler is named, so GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot read the whole store. The file now opens with a note for shopping agents, pointing them at /agents.md and the store's commerce endpoints, and since May 2026 every store serves an /llms.txt and /llms-full.txt that copy /agents.md: instructions for an agent buying on someone's behalf, not a list of the store's pages. Replacing robots.txt means a robots.txt.liquid template, which drops Shopify's managed file — and Shopify's own robots.default_groups still returns an older rule set, with /search disallowed and no note for agents.
robots.txt
# Shopify storefront. Public product, collection, page, blog, policy, cart, and localized HTML is crawlable.
# Agent instructions: https://example.com/agents.md
# UCP discovery: https://example.com/.well-known/ucp
…
User-agent: *
Allow: /
…
Disallow: /admin
Disallow: /cart/
Disallow: /checkout
Disallow: /checkouts/
Disallow: /orders
Disallow: /account
…
Sitemap: https://example.com/sitemap.xmlWhat we read on 48 Shopify sites
Shopify stores of well-known brands, and stores AI answers cite, read on the way an AI crawler reads them: robots.txt first, then the homepage as it arrives, before any JavaScript runs.
44 of 48
serve an /llms.txt
0 of 47
robots.txt files block an AI crawler
36 of 48
have JSON-LD on the homepage
1,378
words on the median homepage before any JavaScript runs
What you can change, and where
The changes a report asks for on the site itself, and llms.txt, which it does not.
Page titles & meta descriptions
Possible: YesWho: You, with stepsWhere: Home page: Online Store → Preferences. Products, collections, pages and posts: 'Search engine listing' → Edit. In bulk, the SEO title and description columns of the product CSV.See what is live: Website metadata checker
JSON-LD schema
Possible: YesWho: You, with stepsWhere: The theme already adds Organization, Product (or ProductGroup) and Article. More goes in the theme's code or a metafield the theme prints; ratings come from the review app.See what is live: Schema markup validator
robots.txt
Possible: YesWho: You, with stepsWhere: A robots.txt.liquid template: Online Store → Themes → Edit code → Add a new template → robots. It replaces the managed file; deleting the template brings it back.See what is live: robots.txt validator
A plain-text file like /llms.txt
Possible: YesWho: You, with stepsWhere: Served for you. To write your own, add templates/agents.md.liquid, which covers /agents.md, /llms.txt and /llms-full.txt. A template sees only the request, so it cannot list the store's products.See what is live: llms.txt validator
New pages (FAQ, comparison, use case)
Possible: YesWho: We write; whoever runs the site puts it upWhere: Online Store → Pages, or a post in the store's blog.A new section on an existing page
Possible: YesWho: We write; whoever runs the site puts it upWhere: The page it goes on, in the normal editor, beside the section it follows.
Getting a Shopify site ready to be cited
- 01
Read the live /robots.txt and /llms.txt
Both are Shopify's until a theme template replaces them. A template copied from an old guide may still disallow /search and leave out the note for agents.
- 02
Give collection pages copy above the grid
Two or three paragraphs on who the collection is for and how to choose within it. A grid of tiles has nothing to quote.
- 03
Make sure reviews reach the HTML
A review app that loads by script is invisible to a crawler that does not run JavaScript. View the page's source and search for a sentence from a review.
- 04
Look at Sales channels → Agentic
It lists the AI channels Shopify sells through — ChatGPT, Copilot and others — and, since Spring '26, the AI queries your products come up for.
- 05
Claim the review profiles the report names
On ecommerce reports the answer often cites a review site or a 'best X' roundup rather than a store. The report lists the profiles you can claim, like Trustpilot, and lists a roundup as an article.
Check it yourself
Free, no account. Each reads what your site serves today, the way the crawlers do.
Why we don’t publish to Shopify
We publish into WordPress and Webflow, not into Shopify. An approved page is copied — rich text for the editor, HTML or Markdown — and pasted in, by you or by the site owner from a link you send them; once either of you gives us its address, we look for it on the site. Listings, reviews and site files are tasks with steps.
Passcite on Shopify
The reportLive
It reads public pages, so it works the same on Shopify as anywhere.
Pages for your approvalLive
Written from the site's own words for the prompts it loses.
Publishing into ShopifyCopied in
Rich text, HTML or Markdown, looked for on the site once it is up.
ProofLive
A week after a page goes live, its prompts are asked again.
Questions
- Does Shopify block AI crawlers?
- No. Shopify's managed robots.txt names no AI crawler, so GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot are allowed. Shopify does rate-limit bots that don't sign their requests, at its network rather than in robots.txt.
- What is the llms.txt Shopify serves?
- Since May 2026 every store serves /llms.txt and /llms-full.txt as copies of a generated /agents.md: instructions for an AI agent shopping on a buyer's behalf — the store's commerce endpoints, its policies and how checkout works. It does not list your pages. A theme template, agents.md.liquid, replaces all three.
- Do I still need to add product schema?
- The theme already writes it. What it does not write is a page that compares — the collection copy or buying guide that answers 'which one for me', which is where most Shopify reports find the gap.
- Do I need to give Passcite access to my Shopify site?
- No. The report reads public pages, the same as any visitor, and nothing we do connects to your Shopify account.
- What does Passcite cost?
- The first report is free on a free account: one brand, 10 prompts, three engines, no card. Plans start at $49 a month for 150 credits — a prompt asked on three engines is one, an article 10.
- WordPressThe platform we publish into. An SEO plugin serves most of what AI crawlers read; the page itself is the gap.
- BigCommerceEvery AI crawler may read the store, slowly: a hidden robots.txt group asks 42 of them to wait ten seconds.
- WixWix writes robots.txt, structured data and an llms.txt, and lets every AI crawler in. Content is the gap.
- SquarespaceAI crawlers allowed, structured data automatic, and one box in Settings → Crawlers that can turn 26 of them away.
See what the engines say about your Shopify site
The prompts it loses, who was named instead, and what the site never answers.