# pottery-store.com — crawler policy
#
# Served as a real file from /srv/pottery-store/app. Before this existed the
# SPA catch-all answered /robots.txt with the app shell, so crawlers got an
# HTML page instead of a policy and sitemap discovery was impossible.
#
# RFC 9309: a named crawler obeys its OWN group and inherits nothing from `*`.
# The search and query-string limits therefore repeat in every group — a fence
# only in `*` fences off nobody who actually generates the load. That was this
# file's state until 2026-08-18, and our own CRAWL-01 failed us for it.
#
# The limits cost no AI visibility: product and category pages stay open, and
# the sitemap lists every product. What is fenced off is catalog search and
# faceted query-string URLs — a combinatorial, uncacheable space that holds
# nothing an agent reads or buys from, and that turns a parallel crawl into an
# outage.

User-agent: GPTBot
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: OAI-SearchBot
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: ChatGPT-User
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: ClaudeBot
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: Claude-User
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: PerplexityBot
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: Google-Extended
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

User-agent: Bingbot
Allow: /
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

# Everyone else, including the crawlers without a group of their own. The
# session- and state-carrying paths are fenced here as well: they are about
# what a page holds, not about who is asking.
User-agent: *
Allow: /
Disallow: /checkout
Disallow: /cart
Disallow: /account
Disallow: /customer
Disallow: /graphql
Disallow: /rest/
Disallow: /search
Disallow: /szukaj
Disallow: /*/szukaj
Disallow: /catalogsearch/
Disallow: /*?

Sitemap: https://pottery-store.com/sitemap.xml