# robots.txt — mygardeningzone.com

# --- Googlebot: allow assets and admin-ajax
User-agent: Googlebot
Allow: /*.js$
Allow: /*.css$
Allow: /wp-admin/admin-ajax.php
Disallow: /wp-admin/
Disallow: /?s=
Disallow: /*?s=
Disallow: /search/
Disallow: /*/feed/
Disallow: /wp-login.php
Disallow: /wp-register.php
Disallow: /trackback/

# --- All other well-behaved bots
User-agent: *
Allow: /wp-admin/admin-ajax.php
Disallow: /wp-admin/
Disallow: /wp-json/
Disallow: /?s=
Disallow: /*?s=
Disallow: /search/
Disallow: /*/feed/
Disallow: /wp-login.php
Disallow: /wp-register.php
Disallow: /trackback/
Crawl-delay: 5

# --- Amazon's search crawler, named explicitly.
# MEASURED 2026-07-25 (two days of origin logs, three sites): Amzn-SearchBot spends
# 40.6% (cookingvibe) / 47.7% (droolrecipes) / 49.2% (recipenames) of its requests on
# /wp-json/ — the REST API — and essentially none on search, feeds, replytocom or
# attachments. So the usual "crawl-waste" Disallow list would have achieved NOTHING here;
# this one stanza removes roughly half of the single largest controllable load on the box,
# and costs zero article visibility because /wp-json/ is not content.
# It must be NAMED: Amazon runs Amzn-SearchBot on its own token and its crawlers are
# reported to ignore the wildcard group, so the block above does not reach it. Amazonbot — the
# token our old blocklist and Cloudflare's managed robots.txt both use — is a DIFFERENT
# crawler that sends us 0 requests/day.
# Articles stay crawlable on purpose: this throttles the waste, not the indexing.
User-agent: Amzn-SearchBot
Allow: /wp-admin/admin-ajax.php
Disallow: /wp-json/
Disallow: /wp-admin/
Disallow: /?s=
Disallow: /*?s=
Disallow: /search/
Disallow: /*/feed/

# --- Aggressive / low-value bots (full block)

User-agent: Leikibot
Disallow: /

User-agent: serpstatbot
Disallow: /

# --- Sitemap
Sitemap: https://mygardeningzone.com/sitemap_index.xml