# Yungle — https://yungle.co
#
# Two things about the format that this file got wrong once each, so they are
# written down here rather than rediscovered:
#
# 1. A BLANK LINE TERMINATES A RECORD. Comments may sit between directives, but
#    an empty line ends the group belonging to a `User-agent`. An earlier version
#    had explanatory paragraphs separated by blank lines between `User-agent: *`
#    and its rules, which orphaned every Disallow — the file parsed cleanly and
#    disallowed nothing at all. Keep the directives below contiguous.
#
# 2. NO BLANKET `Allow: /`. It is redundant (anything not disallowed is allowed)
#    and parsers disagree about precedence: RFC 9309 and Google use longest-match,
#    where it is harmless, but first-match-wins parsers take it as "crawl
#    everything" before reaching any Disallow.
#
# Both were found by running a real robots parser against the file, not by
# reading it. Do that again after editing.
#
# Also deliberate: /t/, /c/ and /r/ are NOT disallowed. Those are
# capability URLs and each page sets `robots: { index: false }` itself, which is
# the correct mechanism — a crawler told not to fetch a page can never see the
# noindex on it, so disallowing them would make indexing more likely, not less.
# They are absent from the sitemap by construction.

User-agent: *
# The signed-in app: not content, and every route needs a session anyway.
Disallow: /dashboard
# API endpoints are not content either — but the OpenAPI document IS. It is the
# machine-readable contract and the most useful thing an agent can find here.
# The exception must precede the rule for first-match parsers, and wins on
# longest-match too, so it is correct under both.
Allow: /api/v1/openapi.json
Disallow: /api/

Sitemap: https://yungle.co/sitemap.xml