# RailRoute — https://www.railroute.in
# Canonical host is https://www.railroute.in; the apex redirects here.

User-agent: *
Allow: /

# API surface: JSON only, nothing to index, and every route is reachable
# from a rendered page anyway.
Disallow: /api/

# Development scratch pages. Also noindex via X-Robots-Tag at the edge.
Disallow: /demo
Disallow: /modal-demo

# Auth callbacks and account screens are per-user and never public.
Disallow: /api/auth/

# The admin panel. Listed for crawlers that honour this file, and protected
# by authentication for the ones that do not — a stranger gets the same 404
# a nonexistent route gives. See src/lib/admin/guard.ts.
Disallow: /admin
Disallow: /admin/
Disallow: /api/admin/

# Search results. Every /search?from=&to= URL is noindex already, and
# Search Console counted 9,873 of them crawled by 2026-10-10 while 21,691
# real pages waited in "Discovered - currently not indexed". The crawl is
# the budget those pages need. Bare /search stays open. The assistant group
# below keeps /search, because those agents fetch it for a person.
Disallow: /search?

# ---------------------------------------------------------------
# Assistants and retrieval crawlers.
#
# Named explicitly rather than left to the wildcard, because several of
# these read only a rule addressed to them and a couple treat the absence
# of one as a reason not to fetch at all. The public pages are open on
# purpose: the timetable pages are published schedule information and the
# search is the product. The API is closed to all of them — it is JSON
# with nothing to cite, and every route behind it is reachable from a
# rendered page.
#
# NOTE THAT THIS PROTECTS NOTHING. robots.txt is advisory and honoured only
# by clients that choose to; `Disallow: /api/` saves crawl budget and is not
# access control. What protects the API is authentication and rate limiting.
#
# Google-Extended is not a crawler. It controls whether pages Googlebot has
# already fetched may be used for Gemini grounding and training, and has no
# effect on Search indexing. Listed for the avoidance of doubt.
#
# THE NAMES BELOW WERE CHECKED AGAINST THE OPERATORS' OWN DOCUMENTATION,
# not copied from an SEO post. As of August 2026:
#
#   OpenAI (developers.openai.com/api/docs/bots)
#     OAI-SearchBot  search indexing for ChatGPT search
#     ChatGPT-User   fetches a page because a person asked about it
#     GPTBot         training
#   Anthropic (support.claude.com, article 8896518)
#     Claude-SearchBot  search indexing
#     Claude-User       fetches a page because a person asked about it
#     ClaudeBot         training
#
# Claude-User WAS MISSING FROM THIS FILE and is the one that matters most of
# the three: it is the agent that fetches a page when somebody asks Claude a
# question this site can answer. Claude-Web was listed and has been removed —
# Anthropic no longer documents it, and a rule addressed to a retired token
# is noise that makes the real ones harder to audit.
#
# THE TRAINING CRAWLERS STAY ALLOWED, which is the policy this file already
# had rather than a new decision. It is recorded here so that changing it is
# deliberate: appearing in AI search results does NOT require allowing
# training, the two are separate crawlers with separate tokens, and the
# search ones above would keep working unchanged if the training ones were
# disallowed tomorrow.
# ---------------------------------------------------------------

User-agent: GPTBot
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
User-agent: ClaudeBot
User-agent: Claude-SearchBot
User-agent: Claude-User
User-agent: PerplexityBot
User-agent: Perplexity-User
User-agent: Google-Extended
User-agent: Applebot-Extended
User-agent: Amazonbot
User-agent: cohere-ai
User-agent: Meta-ExternalAgent
User-agent: CCBot
Allow: /
Disallow: /api/
Disallow: /admin
Disallow: /admin/
Disallow: /api/admin/
Disallow: /demo
Disallow: /modal-demo
Disallow: /map-demo
Crawl-delay: 10

Sitemap: https://www.railroute.in/sitemap.xml

# A curated map of this site for agents, in llms.txt v2 form. Not a
# directive — robots.txt has no such field, and no crawler is expected to
# read this line. It is here because it is where a person looks first.
# https://www.railroute.in/llms.txt