# Content signals, carried over from the live site.
# search:   building a search index and returning links and short excerpts.
# ai-input: feeding content into an AI model at answer time.
# ai-train: training or fine-tuning AI models.

User-agent: *
Content-Signal: search=yes,ai-train=no,use=reference
Allow: /
# One page every ten seconds, so no single crawler can walk the encyclopedia
# faster than we want to pay for. Google ignores this line by design (its rate
# lives in Search Console); Bing, Yandex and most of the smaller crawlers honour
# it, and they are the bulk of the traffic this is aimed at.
Crawl-delay: 10

# The CMS, the API and preview routes. Nothing here is a page.
Disallow: /admin
Disallow: /api
Disallow: /next
# The token reference. Real, but not a page anyone should land on from a search.
Disallow: /design

# Training-only crawlers. Blocked: they take the corpus and send nobody back.
# From Cloudflare's managed list on live (2026-07-29), less the two below.
User-agent: Amazonbot
Disallow: /

User-agent: Bytespider
Disallow: /

User-agent: CCBot
Disallow: /

User-agent: ClaudeBot
Disallow: /

User-agent: CloudflareBrowserRenderingCrawler
Disallow: /

User-agent: GPTBot
Disallow: /

User-agent: meta-externalagent
Disallow: /

# Allowed on purpose (Sidian, 2026-07-30): these feed AI answers that cite and link
# their sources, so an archetype quoted there can still send a reader here.
User-agent: Google-Extended
Allow: /

User-agent: Applebot-Extended
Allow: /

Sitemap: https://mymythos.org/sitemap.xml