# CheckedCare crawler policy: READ and CITE, yes. TRAIN, no.
#
# The product is the pipeline and the current, cited records, not secrecy, so we
# want answer engines to fetch us live and cite us. We do NOT want the data
# absorbed into model weights, where it can be reproduced without a fetch,
# without a citation, and without ever going stale. That disintermediation is the
# one outcome that undercuts the business. So the training crawlers are
# disallowed outright, and every crawler we do allow is told ai-train=no.
#
# Two layers must agree: this file (which honest bots read) and Cloudflare's edge
# "AI bot policies" Training = Disallow toggle, which enforces the same intent for
# bots Cloudflare can fingerprint, including ones that ignore robots.txt. Keep
# both in sync.
#
# Content-Signal (contentsignals.org) declares permitted USE, which is separate
# from crawl ACCESS (Allow / Disallow). It is repeated inside each group because a
# crawler obeys only its single most-specific matching group, so a signal under
# "*" alone would never reach the named bots below.

User-agent: *
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

# --- Search and answer-engine crawlers: allowed to READ and CITE (ai-train=no) ---

# Googlebot drives Search and AI Overviews. Gemini training is opted out
# separately via the Google-Extended disallow further down; disallowing
# Google-Extended does not affect Search or AI Overviews.
User-agent: Googlebot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: Bingbot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: Applebot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

# OpenAI: the search index and live user fetches cite us. GPTBot (training) is
# disallowed below, which affects neither of these.
User-agent: OAI-SearchBot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: ChatGPT-User
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

# Anthropic: live user fetch and the search index cite us. ClaudeBot (the general
# training crawler) is disallowed below.
User-agent: Claude-User
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: Claude-SearchBot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: Claude-Web
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

# Perplexity is an answer engine built on citation.
User-agent: PerplexityBot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: Perplexity-User
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: DuckAssistBot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

# Amazon (Alexa / Rufus) and Meta AI each use one blended crawler for both
# indexing and training, so there is no separate citation bot to allow. We keep
# the read-and-cite access and rely on the ai-train=no signal to decline training.
User-agent: Amazonbot
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

User-agent: meta-externalagent
Content-Signal: search=yes, ai-input=yes, ai-train=no
Allow: /

# --- Training crawlers and training-only opt-out tokens: DISALLOWED ---
# Blocking each of these costs zero citation: either the same company's search or
# live crawler is allowed above, or the bot only feeds a training corpus with no
# answer engine that would ever cite us back.

User-agent: GPTBot
Disallow: /

User-agent: ClaudeBot
Disallow: /

User-agent: anthropic-ai
Disallow: /

User-agent: Google-Extended
Disallow: /

User-agent: Applebot-Extended
Disallow: /

User-agent: CCBot
Disallow: /

User-agent: Bytespider
Disallow: /

# --- Bulk scrapers: DISALLOWED ---
# Not a search or answer engine. Also refused with a 403 in the Worker
# (src/lib/bot-deny.js), since a scraper may ignore this file.

User-agent: ShapBot
Disallow: /

Sitemap: https://checkedcare.com/sitemap-index.xml
Sitemap: https://checkedcare.com/facility-sitemap-index.xml

# A plain-text map of the site for AI assistants (llmstxt.org format). Not a robots
# directive; noted here because this is where a reader looks for such a pointer. It is
# also declared as <link rel="alternate" type="text/plain" href="/llms.txt"> in the head.
# https://checkedcare.com/llms.txt