User-agent: *
Allow: /
Disallow: /dashboard
Disallow: /profile
Disallow: /auth
Disallow: /admin

# Private-by-link surfaces. All three already send noindex, but a crawler that never fetches
# them can't leak a token through a referrer or a cached snippet either. /preview/ is the free
# sample of somebody's own episode: one-visitor, one-token, and nothing anyone else should land
# on from a search result.
Disallow: /t/
Disallow: /embed/
Disallow: /preview/

# Block parameter-based duplicate pages
Disallow: /*?*utm_source=
Disallow: /*?utm_source=
Disallow: /*?*utm_medium=
Disallow: /*?utm_medium=
Disallow: /*?*utm_campaign=
Disallow: /*?utm_campaign=
Disallow: /*?*ref=
Disallow: /*?ref=
Disallow: /*?*fbclid=
Disallow: /*?fbclid=
Disallow: /*?*gclid=
Disallow: /*?gclid=

# Block blog search parameter pages (not real content)
Disallow: /blog?*
Disallow: /news?*

Sitemap: https://podcaststotext.com/sitemap.xml

# Content Signals (https://contentsignals.org/)
# We allow search indexing and AI context use, but opt out of AI training datasets.
Content-Signal: ai-train=no, search=yes, ai-input=yes