# robots.txt for VeloWorkout
#
# The rules below are prefix matches, not path segments: "Disallow: /perform/"
# only blocks URLs that BEGIN with /perform/, which is why /workouts/<id>/perform
# was crawled several hundred times before anyone noticed. Anything that can
# appear deeper in a path needs a leading wildcard. See src/test/robots.test.ts,
# which asserts these patterns against the URLs Search Console actually flagged.

User-agent: *
Allow: /

# Whole sections that are private or non-content.
Disallow: /api/
Disallow: /admin/
Disallow: /analytics/
Disallow: /auth/
Disallow: /coach/
Disallow: /history/
Disallow: /planned-workouts/
Disallow: /profile/
Disallow: /redirect/
Disallow: /save-workout/
Disallow: /settings/
Disallow: /subscription/
Disallow: /test/
Disallow: /debug-*

# App surfaces that hang off a content URL. The workout player and editor exist
# once per workout, so leaving them open put roughly 900 crawlable URLs in front
# of Googlebot that it was never going to index.
Disallow: /perform/
Disallow: /*/perform
Disallow: /*/edit

# The route builder is one page. Every workout page links to it with its own
# ?workoutId=, which turned one app screen into hundreds of near-identical URLs.
# The page itself stays crawlable; only the parameterised copies are blocked.
Disallow: /routes?

# Deliberately still crawlable: /*/opengraph-image. Blocking those would save
# crawl budget but Facebook and LinkedIn read robots.txt too, and a blocked card
# is a broken link preview.

# Sitemap
Sitemap: https://veloworkout.com/sitemap.xml
Sitemap: https://blog.veloworkout.com/sitemap.xml