# ecosmart.hu
#
# Deployed to /var/www/ecosmart.hu/public_html/robots.txt.
# nginx serves this statically: `location = /robots.txt` in ecosmart.hu.conf never reaches PHP,
# so WordPress's virtual robots.txt is not in play and this file is the whole truth.
#
# WHY THIS FILE EXISTS (2026-09-07, operator's ruling on the Desk)
# Until today ecosmart.hu had no robots.txt at all: the origin returned 404 and Cloudflare's
# managed robots.txt was serving in its place, carrying Disallow rules for Google-Extended,
# GPTBot, ClaudeBot, CCBot, Bytespider, Amazonbot, Applebot-Extended and meta-externalagent.
# Google-Extended is the one that cost real visibility: it keeps the site out of Google's AI
# Overviews, which now sit ABOVE the classic results for exactly the "megéri a ...?" questions
# /d-arszabas-kalkulator/ was written to answer. The operator approved unblocking it site-wide.
#
# Cloudflare's toggle is all-or-nothing (Security -> Settings -> Bot traffic -> "Set your
# preference to block training in robots.txt"), so the sequence is: this file lands first, the
# toggle goes off second. While the toggle is still on, Cloudflare appends its managed block
# below whatever is here.
#
# Content signals: search and AI answers yes, model training no. That is a deliberate split, not
# a loosening of everything. "ai-input" is what an assistant uses to ground an answer and cite
# the page, which is the visibility we want; "ai-train" is donating the corpus, which we still
# decline. The reservation below is kept verbatim in substance from the file Cloudflare was
# serving, because dropping it would quietly give up a rights reservation we currently hold.
#
# ANY RESTRICTIONS EXPRESSED VIA CONTENT SIGNALS ARE EXPRESS RESERVATIONS OF RIGHTS UNDER
# ARTICLE 4 OF THE EUROPEAN UNION DIRECTIVE 2019/790 ON COPYRIGHT AND RELATED RIGHTS IN THE
# DIGITAL SINGLE MARKET.
#
# Deliberately minimal otherwise. Cart, checkout and account pages are NOT listed: their slugs
# would have to be guessed, a wrong guess silently blocks a real page, and those pages already
# carry their own noindex. A robots.txt is the one file on the site where an invented line can
# deindex the lot.

User-agent: *
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php

Sitemap: https://ecosmart.hu/sitemap_index.xml

# SEO-tool crawlers, blocked 2026-09-11 (operator's ruling). Measured 2026-09-07 at the origin:
# SemrushBot ~735 and AhrefsBot ~78 requests per ~45 h, ten times every AI crawler combined,
# and nothing comes back: no visitor, no citation, no index. The one reason to allow them
# would be our own marketing using Semrush/Ahrefs to audit this site; the operator confirmed
# nobody does. Remove these four lines if that ever changes, or the audits will show a site
# that refuses to be crawled.
User-agent: SemrushBot
Disallow: /

User-agent: AhrefsBot
Disallow: /