lintpage
get started
~/llms-txt
§ proposed standard · 8 pages

llms.txt, without the guessing.

A markdown file at your site root that hands AI assistants a curated index of your pages. Everyone has an opinion about whether it works. Almost nobody has checked. These pages cover what the spec actually requires, how to serve the file on your stack, and what to do when a tool tells you yours is wrong.

check your llms.txt →generate one →
§ the question everyone actually has

Does llms.txt do anything yet?

On the public evidence available today: no, not measurably. No major AI assistant has documented reading llms.txt to decide what to retrieve or cite, and the case studies claiming otherwise do not describe a mechanism - they describe traffic changing while several other things also changed.

That is worth saying plainly, because the alternative is what most pages on this topic do: describe the format, imply it matters, and let the reader assume the implication is evidence. The file is cheap to add and harmless. Add it because you want a curated index of your site to exist in a machine-readable form, and because if clients do start reading it you will already be there. Do not add it expecting a traffic change this quarter.

What does affect whether AI assistants can see your site today is far more boring: whether their crawlers can fetch your pages at all, whether your content survives without JavaScript, and whether a CDN rule is returning 403 to them. Those are measurable right now. The AI Crawler Checker tests them against your URL.

§ by platform

How to serve it on your stack.

Where the file goes, how that platform serves it, and the specific way it goes wrong there. Two of these pages exist to tell you the file cannot be served at all on that platform, which is the answer nobody else gives you.

01llms.txt on Next.jsTwo ways to serve /llms.txt from Next.js - a static file in public/, or a route handler that generates it from your content. Plus the middleware and rewrite traps that make a correct file return HTML.02llms.txt on WordPressWordPress routes every unknown URL through index.php, which is why an uploaded llms.txt can return your 404 template with a 200 status. Three ways to serve it, and how to tell which one your host allows.03llms.txt on Yoast SEOWhether Yoast generates an llms.txt file, how it interacts with the one you wrote by hand, and how to serve your own without fighting the plugin.04llms.txt on ShopifyShopify does not let you put arbitrary files at the domain root, so /llms.txt is not directly writable. What you can actually do, and which workarounds are worth it.05llms.txt on DrupalThree ways to serve /llms.txt from Drupal, why a file in the docroot can vanish on your next Composer deploy, and the .htaccess rule that blocks it on some hosts.06llms.txt on WebflowWebflow hosting does not serve arbitrary root files, so /llms.txt cannot be uploaded. The options that exist, ranked by how much they actually get you.07llms.txt on SquarespaceSquarespace does not expose the domain root, so there is no way to serve /llms.txt from a Squarespace-hosted site. What that leaves you, stated plainly.
§ diagnostics

When a tool says your file is wrong.

The message a checker printed at you, and what it means - including which rules come from the spec and which are the tool’s own opinion presented as a standard.

msgllms.txt does not follow recommendationsA tool told you your llms.txt does not follow recommendations, without saying which one. Here is every rule the format actually defines, which ones are the spec and which are a tool's opinion, and how to tell them apart.
§ the format

The whole spec, in one block.

This is not a simplification. The format really is this small: one required H1, a recommended summary, and sections of annotated links.

/llms.txt
# Project name                    ← required. Exactly one.

> One sentence on what this is.   ← optional in the spec, expected by every tool

Any free-form prose you want here, with no headings in it.

## Docs                           ← zero or more sections

- [Quickstart](https://example.com/docs/quickstart): What it covers
- [API reference](https://example.com/api): Every endpoint and its shape

## Optional                       ← reserved name: a client short on context
                                    may stop reading here
- [Changelog](https://example.com/changelog): Release history

The spec lives at llmstxt.org. Where our checker asserts something the spec does not, it says so on the finding.

§ faq

Questions, answered.

What is llms.txt?
A proposed standard, specified at llmstxt.org, for a markdown file at your site root that gives AI assistants a short curated index of your most useful pages. It opens with an H1 project name and a one-line blockquote summary, then lists links under H2 sections in the form "- [Title](url): notes". It grants no permissions and blocks nothing - it is a reading list, not a policy file.
Does llms.txt actually do anything?
There is no confirmed evidence that any major AI assistant reads llms.txt to decide what to retrieve or cite today. No operator has documented using it, and nobody publishing "llms.txt boosted our AI traffic" case studies has shown a mechanism. It is cheap to add and costs nothing if it turns out to matter later, which is a fine reason to add one. It is not a reason to believe it is working.
Is llms.txt the same as robots.txt?
No, and conflating them causes real mistakes. robots.txt is an access-control convention crawlers have obeyed for thirty years: it says who may fetch what. llms.txt says nothing about permission - it is a curated index for clients that already have access. You need robots.txt to allow the crawler before an llms.txt can matter to it, which is why a site that blocks GPTBot and publishes an llms.txt has contradicted itself.
Where does the llms.txt file go?
At the root of your domain, at https://yoursite.com/llms.txt, served with a text/plain or text/markdown content type. On most stacks that is one file in a public directory. On hosted site builders it can be impossible - Squarespace and Webflow do not expose the domain root at all, and Shopify reserves it. The platform pages here say which case you are in.
What is llms-full.txt?
The expanded companion: the full text of the pages your llms.txt links to, concatenated into one file, so a client can read your documentation without fetching twenty URLs. It is optional, it is a maintenance commitment, and a stale llms-full.txt is worse than none at all. Add it only if you can generate it as part of your build.

LintPage is not affiliated with llmstxt.org or with any AI assistant operator named on these pages. The format is quoted so you can check your own file against it.

§ the measurable part

Check what AI crawlers can actually see.

An llms.txt cannot help a site that returns 403 to GPTBot or renders its content in the browser. LintPage runs 60 checks against a URL in about 30 seconds. Free, no signup.

run a full scan →