murattunalı.

llms.txt: the short path to making models read your site.

Murat Tunalı PUBLISHED: 2026-08-21 READING: 1′ GEO · Teknik

robots.txt speaks to the crawler; llms.txt speaks to the model. A single Markdown file at the root answers “what is this site, what does it offer?” in the plainest form a model can read.

It is a proposal, not a standard yet — and it costs as much as a text file. For an engine that ignores it, nothing changes; an engine that reads it hears about your site from you — a statement, not a guess. Publishing it from day one carries no risk; waiting earns nothing.

What's inside

A title, a one-sentence definition, the services, the live work. No long marketing copy — short lines, clear enough for models to cite. This site's file is nine lines.

Don't leave your site to the model's guess: tell it in your own words.
  1. # murattunalı.
  2. > Premium butik stüdyo — web tasarım,
  3. > geliştirme ve SEO/GEO danışmanlığı.
  4. ## Hizmetler
  5. - Web tasarım + geliştirme
  6. - SEO + GEO danışmanlığı
  7. ## İşler
  8. - fersantekstil.com · karcem.com.tr
  9. - mikroyazilimdestek.com

Where's the proof

The terminal block on the home page is not a styled copy of this file — it is the file: open /llms.txt and compare. If the two ever drift apart, we see it first — the equality is measured before every release.

All posts

SCHEMA: ARTICLE + PERSON + BREADCRUMBLIST — AUTHOR, DATE AND READING TIME FEED THE SCHEMA