About the Legafo page reader

Legafo is a personal AI app. When a person using it asks about a specific web page, Legafo loads that page to answer their question. This page tells site owners what they will see and how to control it.

What it does

Legafo reads a page only at a person's request: an address they gave, one of their bookmarks, or a result of a web search made for their question. It is not a crawler: it does not follow links or build an index, and we do not keep the pages it reads. It only reads — it does not sign in, fill in or submit forms, or change anything on your site.

How to recognize it

Requests for robots.txt and the first request for a page carry a User-Agent that starts with Legafo/1.0. A web page is then loaded by Cloudflare's browser service, which signs its requests with Web Bot Auth (the Signature-agent header) — see Cloudflare's documentation. A PDF is downloaded by that first request. All requests come from Cloudflare's network.

How to control it

Legafo follows robots.txt. To keep it away from your site, or from part of it, add a group for Legafo:

User-agent: Legafo
Disallow: /

If there is no group for Legafo, the rules for * apply. If a site has no robots.txt, Legafo may read its pages. If robots.txt cannot be fetched — the server returns an error or does not answer — Legafo does not read the page.

Limits

No more than three pages for one question, and a daily cap for each person.

Contact

If Legafo misbehaves on your site, write to support@legafo.com.