agentleFS
Sign inStart free
howseen-aihowseen-ai/howseen-geo-skills/llms-txt/SKILL.md

llms-txt

Skill1 starsChanged 55 days agoMITView on GitHub

Generate a valid llms.txt (and optional llms-full.txt) file for a website so AI models can find and understand its most important pages. Use this whenever the user wants an llms.txt file, asks how to make their site readable by ChatGPT/Claude/Perplexity, mentions llms.txt or "the file AI reads", or is setting up basic AI-visibility hygiene for a domain. Point Claude at a URL and it produces the file plus an honest note on what llms.txt does and does not do.

13 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 13 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep descriptions concrete and specific — they help the model decide what to read.

    line 31
  2. Very specific:

    Offer llms-full.txt: if the site is small enough: same idea, but with the actual cleaned text content of the core pages inlined, so a …

    line 33
  3. Very specific:

    Tell the user how to deploy it: place the file at https://theirdomain.com/llms.txt (root).

    line 35
  4. Distinctive:

    Never present llms.txt as the thing that will get them cited.

    line 43

“Common” is measured: kept by at least 6 in 10 of the 13 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

llms-txt

Generate an llms.txt file: a plain-markdown index at the root of a domain that lists its most important pages so AI systems can orient quickly.

Steps

  1. Fetch the site. Get the homepage and, from its navigation and sitemap (/sitemap.xml if present), identify the pages that actually matter: product, pricing, docs, key blog/guide pages, about/contact. Ignore utility pages (cart, login, legal boilerplate) unless the user asks.
  2. Write llms.txt in this exact structure:
# [Company name]

> [One-sentence description of what the company does and for whom.]

## Core pages
- [Page title](https://full-url): one-line description of what's on it
- [Page title](https://full-url): one-line description

## Guides & resources
- [Page title](https://full-url): one-line description

## About
- [Page title](https://full-url): one-line description

Keep descriptions concrete and specific — they help the model decide what to read. Use absolute URLs.

  1. Offer llms-full.txt if the site is small enough: same idea, but with the actual cleaned text content of the core pages inlined, so a model can ingest everything in one fetch.
  2. Tell the user how to deploy it: place the file at https://theirdomain.com/llms.txt (root). No config needed.

Be honest about what this is

Say this plainly in your output, don't bury it:

llms.txt is hygiene, not a ranking lever. As of 2026, no major AI provider has confirmed they use it to decide who gets cited, and independent analysis (Ahrefs) found the vast majority of published llms.txt files receive near-zero requests from AI crawlers. Ship it — it's cheap, correct, and forward-looking — but don't expect it to move your citations. The real levers are: being genuinely crawlable, having quotable answer-first content, and being mentioned across the web. The other skills in this pack work on those.

Never present llms.txt as the thing that will get them cited. That would be false, and honesty is the whole point.

Built by Raphaël Aubry, founder of Howseen (https://howseen.ai). This skill gives you a one-shot file. Whether AI actually cites you — and whether that changes week to week across ChatGPT, Perplexity, Gemini and Claude — is what Howseen tracks continuously. → https://howseen.ai

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free