llms-full.txt is a single file that holds the full text of your key pages, so a reader can take in the whole thing in one fetch instead of following link after link. It sits next to llms.txt, which is a short index of links. The short file says where things are. The full file carries the words themselves. For a store with a large catalog and a library of guides, the full file is how the guides travel as one document instead of dozens of separate fetches.
Say this plainly, because the status matters. The current proposal text at llmstxt.org defines llms.txt only, and it does not define llms-full.txt. The full file grew up next to the proposal as a companion convention used by documentation platforms, not as part of the proposal itself. No vendor requires it, and no engine is documented as reading it. Build it only after llms.txt exists and is current. It is an extra, not a replacement, and it is not a ticket into an answer.
Where it goes wrong
A full-text file fails in a few specific ways, and most of them come from treating it as a dump instead of a curated document.
- The file holds the whole catalog. Every product page, pasted in full, turns the file into a phone book. A reader given the whole catalog in one fetch has to wade through it to find the guides, and the file stops being useful.
- The file is generated once and never rebuilt. Products change, guides get rewritten, prices move, and the full file still carries the old words. A stale full file teaches the reader the old store.
- The file carries prices. A price printed in a full-text file goes stale the day the price changes, and the file is rebuilt less often than the product page. Prices belong on the product page and in the structured data, not in a static text file.
- The file duplicates the short file instead of complementing it. If llms-full.txt is just llms.txt with the same links and no text, it adds nothing. If it holds the text, the short file should stay the map that points at it.
- The file is too large to be useful. A full file that runs to millions of words cannot be read in one pass, and no reader benefits from a document nobody finishes. A file that big needs to be split by section, not shipped whole.
- The file is published, but nobody checks it loads. A full file at the wrong address, behind a redirect, or served with the wrong status is a file that does not exist to a reader. Publish it, then open it in a browser and confirm it loads.
The steps
Step 1. Finish llms.txt first. The marketer confirms the short file exists at the site root, is current, and follows the proposal format: an H1 with the brand name, a blockquote summary, and H2 sections of links with one-line descriptions. Do not start the full file until the short file is right. The full file is built from the pages the short file points at, so a wrong short file builds a wrong full file. Read our llms.txt guide for the file format and a complete sample.
Step 2. Choose what goes in the full file. The marketer picks the pages whose full text earns a place: the buying guides, the how-to pages, the size and fit guides, the care instructions, and the policy pages shoppers ask about. Leave out individual product pages, sale pages that will expire, anything behind a login, and the cart and checkout pages. A large catalog does not mean a large full file. The catalog is summarized in the short file through its collections. The full file carries the guides, not the products. If the list of chosen pages runs past what a reader can use in one document, split it by section instead of shrinking the guides.
Step 3. Build the file from the live pages. The developer generates the file from the current live pages, not from a draft or a copy. Each page becomes one section in the file, in the same order as the short file lists them, with the page title as the heading and the full text beneath it. Strip the navigation, the cookie banner, the ads, and the layout, and keep the words, the headings, and the links that matter. Confirm the text in the file matches the live page word for word on the facts: the steps, the measurements, the limits. A guide that says one thing on the page and another in the file teaches the reader to distrust both.
Step 4. Keep prices out of it. The marketer checks the full file for any price, sale price, or discount code, and removes each one, replacing it with a link to the product page where the current price lives. Prices change more often than the full file is rebuilt, so a price in the file is a stale price waiting to happen. The same rule holds for stock status. State what a product is and how to use it, and let the product page state what it costs and whether it is in stock. The structured data on the product page is where price and availability belong.
Step 5. Publish it at the site root and link it from llms.txt. The developer publishes the file at the site root, next to llms.txt, and confirms it loads in a browser at your store domain followed by llms-full.txt. Then the marketer adds one line to llms.txt pointing at the full file, so a reader that starts from the short file can find the full one. The file must serve as plain text over HTTPS with a 200 status. There is no submission step, no search console entry, and no ping endpoint. Publish it, verify it loads, and re-fetch it after you change it.
Step 6. Rebuild it on the same routine as the guides. The marketer and the developer rebuild the full file whenever a guide in it changes, a guide is added, or a guide is removed. A file that lags the guides teaches the reader the old guides. Build the rebuild into the same routine that publishes a guide, so the file stays current without a separate project. Check the file size each time it is rebuilt. If it has grown past what a reader can use in one document, split it by section and link the sections from llms.txt instead of shipping one oversized file.
The machine-readable piece
This is the llms.txt index for a store with a large catalog, following the proposal format. It lists the collections and the guides, and it points at the full file. Replace the placeholders with your own pages.
# Example Brew Supply
> Example Brew Supply sells home brewing kits, ingredients, and equipment, shipping across the country. The shop publishes brewing guides and runs a recipe help line.
## Shop
- [Brew kits](https://[yourdomain.com]/collections/brew-kits): Starter and advanced kits, listed with batch size and included equipment.
- [Ingredients](https://[yourdomain.com]/collections/ingredients): Malt, hops, and yeast, listed with style and origin.
- [Equipment](https://[yourdomain.com]/collections/equipment): Kettles, fermenters, and kegs, listed with capacity and material.
## Guides
- [First brew day](https://[yourdomain.com]/pages/first-brew-day): The full first brew day, step by step, with timings and temperatures.
- [Cleaning and sanitizing](https://[yourdomain.com]/pages/cleaning-sanitizing): Cleaning and sanitizing steps for every piece of equipment.
## Full text
- [Full guide text](https://[yourdomain.com]/llms-full.txt): The full text of the guides above in one file.
And this is the shape of the full file itself. Each page becomes one section, in the same order as the short file lists them, with the page title as the heading and the full text beneath it. This excerpt shows two pages. A real file holds every chosen guide, in order, with no navigation, no banner, and no layout.
# Example Brew Supply guides
## First brew day
Example Brew Supply walks a first-time brewer through a full brew day, from heating the water to pitching the yeast. The guide gives the steps in order, the temperatures to hit, and the mistakes to avoid. Prices and stock are not in this file. See the product pages for current prices.
## Cleaning and sanitizing
Example Brew Supply explains how to clean and sanitize every piece of equipment before and after a brew day. The guide lists the steps in order and states which cleaners suit which materials.
Check both files after every rebuild. Open your domain followed by llms.txt and confirm the links load. Open your domain followed by llms-full.txt and confirm the full text loads and matches the live guides. If a guide changed and the full file did not, rebuild it before you rely on it. Do not add a field, a page, or a value you cannot verify against the live site.
What it looks like when it works
On the page, nothing looks different to a human reader. These are files for machines, not pages for shoppers. The change is under the surface: the short file maps the store in a page of links, and the full file carries the guides as one document, in the same order, with the same words as the live pages. A reader that starts from the short file can follow one link and read the whole guide library without fetching dozens of pages. A shopper who asks an assistant a how-to question gets an answer drawn from the current guides, not from a stale copy.
Say you run a home brewing supplies store with a large catalog, selling kits, ingredients, and equipment, plus a library of brewing guides. The marketer finishes llms.txt first, listing the collections and the guides with one-line descriptions. Then the marketer picks the guides for the full file: the first brew day, cleaning and sanitizing, choosing yeast, and kegging, leaving out the product pages and the sale pages. The developer builds the full file from the live guides, strips the navigation and banners, and publishes it at the site root. The marketer checks the file, finds a price printed in one guide, and replaces it with a link to the product page. The developer confirms both files load in a browser. When a guide is rewritten, the full file is rebuilt the same day. A shopper asks an assistant how to sanitize a fermenter before a first brew. The assistant can read the guide in full, and it cites the shop with the steps as written.
The fundamentals
These mechanics are shared across every engine and covered in full in their own guides:
- •The llms.txt format and a complete sample, see our guide to writing a good llms.txt file
- •The markup AI systems read to verify facts, see our product schema guide
- •The complete crawler rule reference, see our robots.txt guide for AI search
What this guide does not cover
This guide covers adding llms-full.txt for a store with a large catalog and a guide library. It does not cover the full technical audit that checks every crawler and every page, the transaction layer that lets an agent complete a purchase, or the per-engine playbooks for getting each assistant to find you. It does not cover keeping prices and availability accurate, which belong on the product page, in the structured data, and in the feed. Those are separate jobs, and this guide is the map, not the whole trip.
Check your own store now
Run the free scan. One click, no signup, and it shows whether AI crawlers can read your store today: aizii.ai/aeo-scan
Run the free scanSources
- llms.txt proposal (specific): https://llmstxt.org
- Google Search Central, Optimizing your website for generative AI features on Google Search (specific): https://developers.google.com/search/docs/fundamentals/ai-optimization-guide
- Google Search Central, Product structured data (specific): https://developers.google.com/search/docs/appearance/structured-data/product
- Schema.org, Product type (specific): https://schema.org/Product
- Schema.org, Offer type (specific): https://schema.org/Offer
Note: links marked specific point to the page that supports the claim in this guide. Links marked general docs point to the vendor's general docs section for that topic, not to a page that proves every step. Check the vendor docs for the current path before relying on a menu name or click path in this guide.
