Shopify now natively generates an llms.txt file and an agents.md file for every store by default, a platform-level change that replaced the redirect workarounds merchants used before it. Beyond that, the practical levers for Shopify AI search visibility are robots.txt rules for named AI crawlers, structured data, and stable URLs. Several emerging agent protocols simply have no real implementation path on Shopify’s hosted infrastructure yet, and this guide skips those rather than describing theory with nothing to act on.
This guide covers AI and agent visibility specifically. For the technical SEO foundation underneath it, robots.txt, sitemaps, canonicals, and structured data, see our Shopify SEO guide.
Shopify AI Search Visibility Starts With llms.txt and agents.md
This is the single biggest recent change in this space. Per Shopify’s developer changelog, every store now includes a default agents.md file accessible at /agents.md, and the paths /llms.txt and /llms-full.txt point to that same content by default. This rolled out platform-wide in 2026 and overrides any previous redirect-based or app-based llms.txt workaround automatically, since Shopify’s native route now takes precedence.
The default generated file is intentionally minimal: links to your All collection page and search page, basic store contact information, and a section aimed at agents and developers referencing Shopify’s own platform-level discovery links. It is functional but generic, since Shopify has no way to know which collections, FAQs, or policies matter most for your specific store.
How to Customize It
Shopify’s official supported method is a theme template, following the same pattern as robots.txt.liquid. Under Online Store > Themes > Edit code, add a new template for llms.txt, llms-full.txt, or agents.md as needed. One important difference from robots.txt: a custom template fully replaces Shopify’s generated output rather than merging with it, so your custom version needs to include everything you want present, not just your additions.
Before you overwrite anything, fetch your current live file at yourdomain.com/llms.txt and save a copy. Preserve the agents and developers section from Shopify’s default, since it links to Shopify’s own agent and discovery infrastructure that supports agentic commerce features. Add your own curated links on top of it: your best-performing collection pages, a FAQ or policy page, and five to ten of your strongest blog posts if you publish content, not your blog index page.
Keep the file focused. A curated executive summary in the 30 to 60 link range, under roughly 50KB, is more useful to an AI system than a dump of every URL on your site, since that duplicates what your sitemap already does without adding any curation value.
Should You Customize It at All?
Given how recently this rolled out, opinions differ on urgency. The most balanced take, from Ilana Davis’s analysis of the native rollout, is that customizing is not required and there is no confirmed evidence yet that any AI search engine actually weights llms.txt content in its answers. It will not make or break your SEO either way. If you have a large catalog, complex buying criteria, or important guides a generic file will never surface, customizing is low-risk and plausibly helpful. If your store is small and simple, the default is a reasonable place to leave things for now.
robots.txt Rules for AI Crawlers
Shopify AI search visibility also depends on which crawlers you actually let through. Major AI companies publish named user-agents for their crawlers, GPTBot for OpenAI, ClaudeBot for Anthropic, PerplexityBot for Perplexity, and others. Since Shopify now allows robots.txt.liquid customization on every plan, you can add explicit allow or disallow rules for any of these by name, the same way you would block or allow a traditional search crawler.
This is a real decision with two reasonable answers depending on your priorities. Allowing AI crawlers means your product and content pages can be cited in AI-generated answers and potentially surfaced to AI shopping agents. Blocking them means your content cannot be used to train or ground future AI responses, a position some merchants take over data control concerns. There is no universally correct choice. Decide deliberately rather than leaving it to whichever default your theme or an app happened to set.
Content Signals in robots.txt
This is an emerging, still-draft proposal that lets a site declare more granular permissions than a flat allow or disallow, specifically whether a crawler may search, ingest, or train on your content. Since Shopify’s robots.txt.liquid template accepts plain text rules in addition to its built-in Liquid objects, adding a Content-Signal directive is technically just adding a line of text to the file you can already edit.
Treat this as genuinely optional and forward-looking. Adoption among AI crawlers is inconsistent today, so a Content-Signal directive may simply be ignored by most bots that fetch your file right now. It costs nothing to add if you are already editing robots.txt.liquid for other reasons, but it is not worth a dedicated trip to your theme code on its own yet.
IndexNow for AI-Adjacent Search
IndexNow pushes instant update notifications to participating engines, Bing chief among them, rather than waiting for a scheduled crawl. This matters for AI visibility specifically because Bing’s index feeds Microsoft Copilot, and several AI answer engines lean on Bing’s index as a data source rather than crawling the entire web independently.
Shopify has no native IndexNow integration, and the protocol needs a verification key file at your domain root, which Shopify does not expose directly. The realistic options are a Shopify Files plus URL Redirect setup for the key, or one of several IndexNow apps available on the Shopify App Store that automate key hosting and submission. Google does not participate in IndexNow regardless of which method you use, so this lever specifically helps Bing-adjacent AI surfaces, not Google’s AI Overviews.
Stable URLs as an Agent Trust Signal
An AI agent or answer engine that cites your store is implicitly trusting that the link it surfaces will keep working. A broken citation looks bad for the agent, not just for you, which is part of why stable URLs matter more in an agentic context than they did in pure search.
The practical discipline is the same one that matters for regular SEO: never delete a URL without a redirect, avoid restructuring collection or product handles casually, and treat every published URL as a long-term commitment once an external system, human or AI, might have linked to it.
Structured Data for Agents
The JSON-LD structured data covered in our main Shopify SEO guide does double duty here. AI agents lean on the same schema.org markup search engines use to extract typed facts about a page, price, availability, return policy, review scores, rather than parsing unstructured prose to guess at them.
A product page with complete, valid Product schema gives an agent or answer engine a confident, structured fact to cite. A page with missing or broken schema forces the agent to infer from text, which increases the odds it gets a detail wrong or skips citing your page altogether in favor of a competitor with cleaner markup.
Machine-Readable Formats Shopify Already Supports
Shopify’s blog feature includes a native Atom feed at /blogs/your-blog-handle.atom, which any feed reader or agent can consume directly without scraping rendered HTML. This is a real, working feature with no setup required beyond having a blog at all.
For richer structured product data access, Shopify’s Storefront API offers a GraphQL endpoint that returns typed product, collection, and inventory data, intended primarily for apps and custom storefronts rather than general AI crawler consumption. It is worth knowing this exists if you are evaluating a custom integration, though it is not a direct substitute for the public-facing signals, llms.txt, structured data, and sitemaps, that general-purpose AI crawlers actually rely on today. If you are still deciding whether Shopify is the right platform for your store at all, our how to start an eCommerce store guide covers that bigger decision first.
What Shopify Does Not Support Yet
A few items worth naming honestly rather than ignoring. Shopify does not give merchants a way to set custom HTTP response headers, such as Link headers for resource discovery, since theme code controls page HTML through Liquid, not the raw HTTP response. There is also no native support for a well-known Agent Skills discovery file, and no realistic merchant-level workaround for it on hosted Shopify infrastructure today.
This is also why this guide skips protocols like A2A agent cards, WebMCP, NLWeb, DNS-based agent discovery, and similar emerging specifications outright. They have no real implementation path on Shopify as it stands, and describing them in detail here would offer nothing actionable. If that changes as Shopify’s own agentic commerce features mature, an update to this guide will reflect it.
FAQ’s
Does Shopify have an llms.txt file by default?
Yes. Every Shopify store now generates an agents.md file by default, with /llms.txt and /llms-full.txt pointing to the same content. This rolled out platform-wide and replaced any previous redirect-based workaround automatically.
How do I customize Shopify’s llms.txt file?
Add a theme template under Online Store > Themes > Edit code for llms.txt, llms-full.txt, or agents.md. A custom template fully replaces Shopify’s generated content rather than merging with it, so include everything you want present, including the agent and developer links from the original default.
Should every Shopify store customize llms.txt?
Not necessarily. There is no confirmed evidence yet that AI search engines weight llms.txt content in their answers. It is low-risk to customize for larger or more complex stores, but the default file is a reasonable choice for a small, simple store for now.
Can I block AI crawlers like GPTBot on Shopify?
Yes. Robots.txt customization is available on every Shopify plan, and you can add named allow or disallow rules for GPTBot, ClaudeBot, PerplexityBot, and other AI crawlers the same way you would for a traditional search bot.
Does Shopify support IndexNow for faster AI-adjacent indexing?
Not natively. You need a Shopify Files plus URL Redirect setup for the verification key, or a third-party IndexNow app. This helps visibility on Bing-fed AI surfaces specifically, since Google does not participate in the IndexNow protocol.
What AI agent protocols does Shopify not support?
Custom HTTP Link headers, Agent Skills discovery files, A2A agent cards, WebMCP, NLWeb, and DNS-based agent discovery all lack a real implementation path on Shopify’s hosted infrastructure today. There is no practical merchant-level workaround for these on the platform as it currently stands.
Why does structured data matter for AI visibility specifically?
AI agents and answer engines rely on the same schema.org JSON-LD markup search engines use to extract structured facts like price, availability, and reviews. This is one of the most overlooked levers for Shopify AI search visibility, since missing or broken schema forces an agent to infer details from prose, increasing the risk of inaccurate citations or being skipped altogether in favor of a competitor with cleaner markup.