# Sanovia production — marketing + public booking crawlable
# NOTE: this file is the PROD robots. staging.sanovia.app never serves it.
# Since 2026-09-29 staging is behind a password (deploy/Caddyfile
# @stagingAuth): every staging page, /robots.txt included, answers 401 to
# anyone without it, and a 401 is never indexed. With the password switched
# off (deploy/staging-basic-auth.sh off), Caddy's own override answers
# instead: an allow-all body («User-agent: *» + «Allow: /», no Sitemap) plus
# X-Robots-Tag: noindex on every staging response. Crawling is allowed there
# ON PURPOSE (SEO fix 1, 2026-09-25): a crawler that may not fetch a staging
# URL never sees its noindex, so pages it already indexed stay indexed.
#
# AI crawlers and agents (GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot,
# Claude-User, PerplexityBot, Amazonbot, meta-externalagent, MistralAI-User,
# CCBot, …) are WELCOME on the same terms as everyone else: the `*` group
# below applies to them too. A curated, token-efficient site summary for
# agents and models lives at /llms.txt (full content: /llms-full.txt).
User-agent: *
Allow: /

Disallow: /login
Disallow: /register
Disallow: /register/
Disallow: /clinic/login
Disallow: /admin
Disallow: /admin/
# The public MCP connector for AI assistants that /llms.txt describes
# (https://sanovia.app/api/v1/mcp). RFC 9309 longest match: this Allow beats the /api/
# Disallow below for /api/v1/mcp only; the rest of /api/ stays blocked.
Allow: /api/v1/mcp
Disallow: /api/
Disallow: /doctor
Disallow: /doctor/
Disallow: /patient
Disallow: /patient/
Disallow: /clinic
Disallow: /clinic/
Disallow: /staff
Disallow: /staff/

Sitemap: https://sanovia.app/sitemap.xml