# Yafa Café — 1785 Portage Ave, Winnipeg, MB R3J 0E8
# https://www.yafacafe.ca/
#
# NOTE ON llms.txt
#   /llms.txt is crawlable by everyone, including Google and Bing. It is kept
#   OUT of the search index with an `X-Robots-Tag: noindex` response header
#   (see the `_headers` file), not with a Disallow rule. Blocking it here would
#   stop search engines from ever reading that noindex header, which is the
#   opposite of what we want. AI agents read the file freely either way.

# ============ Main directives for all bots ============
User-agent: *
Allow: /
Allow: /assets/
Allow: /llms.txt
Allow: /sitemap.xml

# Runtime-injected header/footer fragments. Crawlable so search engines can
# render the site navigation, but noindex'd via _headers so the bare fragments
# never surface as their own results.
Allow: /components/

# Query-string variants of real pages — duplicate content, no crawl value.
Disallow: /*?*author=
Disallow: /*?*tag=
Disallow: /*?*month=
Disallow: /*?*view=
Disallow: /*?*format=
Disallow: /*?*utm_
Disallow: /*?*fbclid=
Disallow: /*?*gclid=

# Paths that must never be exposed if they are ever added.
Disallow: /api/
Disallow: /config/
Disallow: /account/
Disallow: /search/

# ============ AI / LLM crawlers ============
# Explicitly welcome to read the site and llms.txt. Legal pages are withheld
# because they carry no useful information about the restaurant.
User-agent: GPTBot
User-agent: ChatGPT-User
User-agent: OAI-SearchBot
User-agent: ClaudeBot
User-agent: Claude-User
User-agent: Claude-SearchBot
User-agent: anthropic-ai
User-agent: PerplexityBot
User-agent: Perplexity-User
User-agent: Google-Extended
User-agent: Applebot
User-agent: Applebot-Extended
User-agent: CCBot
User-agent: cohere-ai
User-agent: Meta-ExternalAgent
User-agent: meta-externalagent
User-agent: FacebookBot
User-agent: Amazonbot
User-agent: Bytespider
User-agent: DuckAssistBot
User-agent: MistralAI-User
Allow: /
Allow: /llms.txt
Allow: /assets/
Allow: /components/
Disallow: /privacy-policy.html
Disallow: /terms-of-use.html

# ============ Ads bots ============
User-agent: AdsBot-Google
User-agent: AdsBot-Google-Mobile
User-agent: AdsBot-Google-Mobile-Apps
Allow: /

# ============ Crawl delay for heavy bots ============
User-agent: Baiduspider
Crawl-delay: 10

User-agent: YandexBot
Crawl-delay: 10

# ============ Sitemap ============
Sitemap: https://www.yafacafe.ca/sitemap.xml