Skip to main content

Fix llms.txt format errors so AI parsers accept it

llms.txt must start with H1 (site name), then a blockquote summary, then H2 sections of bullet links.

Worth up to 3 points

Why it matters

The llms.txt spec is intentionally strict so parsers can extract structured data without an LLM. A free-form Markdown file at /llms.txt may still be read, but won't get the boost of being a recognized spec-compliant file.

How to fix it

  1. 1. Start with H1, then blockquote

    The H1 is your site/product name. The blockquote is a one-line description.

    # Acme
    
    > Invoicing software for freelancers.
  2. 2. Group links under H2 sections

    Each H2 is a category (Docs, Product, Blog). Each bullet is `[Title](absolute-url): one-line description.`

    ## Docs
    
    - [Getting started](https://acme.com/docs/start): 60-second setup.
    - [API](https://acme.com/docs/api): REST endpoints.
  3. 3. Use absolute URLs

    Relative URLs break when llms.txt is fetched out of context. Always use https://yourdomain.com/path.

FAQ

Can I add prose between sections?
Light prose is fine, but every link should still be in a bullet under an H2 — that's what parsers extract.

See your own score

Run a free Crawlable scan to find every check that needs fixing on your site — not just this one.

Check my site

More guides