Jobs to be done · Fix · SEO · Dev

How to generate an llms.txt for your Shopify store.

You read about llms.txt and want a correct one without hand-writing it — and without discovering in two months that it still lists a discontinued product and none of your new collections. The setup is a one-time job; the maintenance is the part that kills hand-rolled files.

Quick answer

Paste your storefront URL into the free llms.txt generator — a valid draft in about 30 seconds. Host it via Shopify Files plus a URL redirect (recipe below), validate, and set a regeneration cadence. Paid plans keep the file current as the catalog changes. Format details in the llms.txt glossary entry.

The slow way: an afternoon of copy-paste, then rot

The manual version: read the llms.txt proposal to learn the format, then build the file by hand — copy URLs for your top collections, flagship products, guides, and policies out of Shopify admin, and write a one-sentence description for each. Then solve hosting, because Shopify has no root-file upload: the workable path is uploading to Files and redirecting /llms.txt to the CDN URL. Test in a private window. Done — an afternoon if you've done it before, most of a day if not.

The real problem arrives later. Two months on, someone asks if the file is current. It isn't: three new collections launched, a hero SKU was discontinued, a policy changed. Nobody owns the update, so the map now misleads the crawlers it was meant to help. That is the recurring pattern with hand-maintained llms.txt files — not wrong at birth, wrong by quarter two.

The eCommerce Insights way

Generate the draft. The free llms.txt generator reads your sitemap, collection structure, and homepage links, picks the 40–80 URLs that matter for AI crawlers, and drafts a description per entry from your metas and schema.
Tighten the descriptions that matter. The drafts are accurate but generic. Spend ten minutes on hero products and top collections so they match how the brand talks — specific, no superlatives. "650-fill down, waterproof shell, unisex sizing, $485" beats "industry-leading outerwear."
Host via Files plus redirect. Shopify admin: Settings → Files → upload; copy the CDN URL; Online Store → Navigation → URL Redirects → redirect /llms.txt to it. Stores fronted by Cloudflare or an edge worker can serve the path directly instead.
Validate the live file. Plain-text content type, no theme wrapper, no redirect chain, every listed URL returning 200. The generator's built-in validator runs the checks in one pass — also usable on a file you wrote by hand (see the validator).
Set the cadence. Manually: regenerate monthly. On a paid plan, eCommerce Insights regenerates on your refresh schedule, watches crawler-log fetches for GPTBot, PerplexityBot, and ClaudeBot, and flags staleness before the file misleads anyone. The Shopify-specific quirks are covered in depth in the llms.txt for Shopify guide.

What "good" looks like

Live /llms.txt returning 200, plain text, no theme wrapperyes

Entries: homepage, top collections, hero PDPs, guides, policies40–80

Descriptions specific, superlative-free, Wikipedia-tonedyes

AI bots fetching weekly (GPTBot, PerplexityBot, ClaudeBot baseline)3+

Fetch counts from eCommerce Insights crawler-log observations as of early 2026, illustrative. The file is one signal among several — pair it with admitted crawlers (robots.txt job) and complete schema for the full crawl surface.

Ask AI about this job

Have your favorite AI engine apply this walkthrough to your store.

Frequently asked questions

Does Shopify natively support serving llms.txt?

Not directly — merchants cannot upload arbitrary files to the domain root. The standard workaround: upload to Settings → Files, copy the Shopify CDN URL, and add a URL redirect from /llms.txt to it under Online Store → Navigation → URL Redirects. Stores with Cloudflare or an edge worker in front can serve it more cleanly at the edge.

What should go in an llms.txt for a Shopify store?

A one-paragraph brand summary, then 40–80 entries covering the homepage, top collections, flagship PDPs, guides, and policy pages — each with a URL and a one-sentence description written like a Wikipedia editor would. No draft or archived products, no marketing superlatives. "650-fill down parka, waterproof shell, unisex, $485" beats "our industry-leading outerwear."

Which AI crawlers actually fetch llms.txt?

In eCommerce Insights crawler-log observations across Shopify stores as of early 2026: GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot, and Google-Extended fetch the file where one is published — typically a handful of fetches per week per bot on a mid-size store (illustrative). Adoption of the standard is advisory, not binding; the file helps where it exists and costs nothing where ignored.

How often should the file be regenerated?

After significant catalog changes — a new collection, a discontinued hero SKU, a policy rewrite. Monthly is a sensible manual baseline. The paid product regenerates automatically on your refresh cadence and flags staleness, which is the real failure mode: hand-maintained files don't die, they quietly rot.

Does llms.txt replace schema or robots.txt?

No — the three do different jobs. robots.txt grants or denies crawler access; Product JSON-LD makes each page parseable; llms.txt is a curated map that tells language models what matters on the site. A store needs all three — see fix product schema and fix robots.txt.

Related jobs

Fix

A correct llms.txt in 30 seconds.

Generated from your live store, validated, and kept current on paid plans.

Generate my llms.txt Read the Shopify guide