llms.txt Generator

Turn your sitemap into a spec-compliant llms.txt in one pass. Edit every line, copy it, publish it at your root.

An llms.txt file is a markdown map of your site for AI agents: a name, a one-line summary, and sections of links with short descriptions. Writing it by hand for a site with a few hundred pages is tedious, so this tool reads your sitemap and drafts it for you. Nothing is stored. The output is yours to edit and publish. At Layer3Labs, we ship one on this site and on the properties we run, which is where the section rules below come from.

Generate Your llms.txt

What the Generator Builds

The format follows the proposal at llmstxt.org, which Jeremy Howard published in September 2024. The spec calls the H1 “the only required section”. The rest follows the layout the spec describes.

  1. H1: your site or project name, taken from your homepage title and editable.
  2. Blockquote: a “short summary of the project, containing key information,” seeded from your homepage meta description.
  3. H2 sections: Docs, Guides, Blog, Comparisons, Tools, Pricing, Company, and Pages, assigned from each URL's path. Each holds - [Title](url): description lines.
  4. Optional: the spec reserves an “Optional” H2 for “secondary information: links an agent can skip when a shorter context is needed.” Untick pages you want out, or move them there by hand.

How to Install It

Copy the output into a file named llms.txt and put it where your site serves static files: the public/ folder on Astro or Next.js, the web root on a plain host, or the static-files location your platform exposes. Load yourdomain.com/llms.txt in a browser. It should return the plain text with a 200 status. Keep it “small enough to fit in context” and put detail “behind the links,” as the spec puts it. Regenerate it when you add a section. You do not need to touch it for every post.

On WordPress: there is no built-in way to serve a root text file from the admin. Upload it over SFTP or through your host's file manager, and check that a caching or security plugin is not rewriting the path. Sites built as files (Astro, Next.js, a Git-based CMS) ship it from the repo, which is one of the reasons those stacks are easier for AI agents to work with; see the best CMS to use with AI.

Does llms.txt Work?

It is a proposal, published by Jeremy Howard in September 2024. The spec page makes no claim about ranking or citations, so treat it as a low-cost map for agents rather than a ranking lever. It costs a few minutes and removes the guesswork for any agent trying to understand your site, which is why we publish one on this site and the properties we operate. The llms.txt explainer walks through what the major engines have said and what belongs in the file.

Frequently Asked Questions

  • Enter your site URL above. The tool reads your sitemap, pulls the title, meta description, and first heading from up to 60 pages, groups them by section, and writes the file in the format the spec describes: an H1, a blockquote summary, then H2 sections of markdown links. Edit any line, copy it, and save it as llms.txt at your site root.
  • The spec at llmstxt.org requires only an H1 with the site or project name. After that it expects a blockquote with a short summary, optional detail paragraphs, and zero or more H2 sections that each hold a list of markdown links with an optional one-line description. An "Optional" section holds links an agent can skip when it needs a shorter context.
  • llms.txt is the index: short, linked, small enough to fit in a model's context. llms-full.txt is a community convention that the spec text does not define: the full page content concatenated into one file. This tool produces the index. The full file is only worth building once the index exists and you know which pages matter.
  • It is a proposal from Jeremy Howard, published in September 2024. The spec page makes no ranking or citation claim, so publish it for the agents that read it rather than for a lift. It costs almost nothing to publish and it is the cleanest way to hand an agent a map of your site, which is why we publish one on the sites we operate. Our llms.txt explainer at /guides/llms-txt-explained goes through the evidence.
  • Because the file is meant to be small enough to fit in context, and because reading pages takes time. If your site is larger, paste the URLs that matter most, or run the tool once per section and merge the results under separate H2 headings.
  • No. The crawl runs on our server for the duration of the request, returns the page titles and descriptions to your browser, and keeps nothing.

Want Your Whole Site Readable by AI?

llms.txt points agents at your pages. The bigger win is content stored as files an agent can read and edit. We audit what you have and plan the move.

Book a Consultation