User-agent: *
Allow: /
Disallow: /blog/drafts/
# Sprint 0 (Workshop) — private auth-gated landing. The worker
# returns 404 for unauthenticated visitors, but defense-in-depth:
# tell crawlers explicitly not to follow these paths even if a
# stale link leaks somewhere.
Disallow: /workbench/
Disallow: /es/workbench/
# Phase 3 (Workshop) — operator account page. Same posture as
# /workbench/: 404-gated server-side, noindex in HTML, and disallowed
# here. /account/ never appears in any sitemap.
Disallow: /account/
Disallow: /es/account/
# Phase 3 — destructive-action confirmation surface. The token in
# the URL is the only thing that makes the page renderable, but
# there is no value to a crawler discovering the path.
Disallow: /api/auth/account-delete-confirm
# Phase G.1 (Growth) — defense-in-depth for surfaces that should
# never appear in search. /admin/* gates server-side via
# _requireAdminSession; /sign-in/ is the magic-link landing;
# /es/blog/drafts/ mirrors the EN drafts disallow above.
Disallow: /admin/
Disallow: /sign-in/
Disallow: /es/sign-in/
Disallow: /es/blog/drafts/

# Phase 7 (launch plan, Part X — AI search posture).
#
# Explicit allows for the AI-search crawlers the plan calls out.
# Without per-agent stanzas, some of these crawlers (notably
# GPTBot and Google-Extended) read the wildcard above as opt-in
# but still benefit from a named allow for citation traceability;
# others (ClaudeBot, PerplexityBot) use named allows to enable
# real-time fetch when a user asks a question that maps to one of
# our pages. The full-body citation corpus lives at /llms-full.txt
# (and /es/llms-full.txt) — see /llms.txt for the map.

User-agent: GPTBot
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: ClaudeBot
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: anthropic-ai
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: Claude-Web
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: PerplexityBot
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: Google-Extended
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: Bytespider
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

User-agent: Applebot-Extended
Allow: /
Disallow: /blog/drafts/
Disallow: /workbench/
Disallow: /account/
Disallow: /admin/
Disallow: /sign-in/

# Aggressive scrapers we explicitly opt out of. These do not power
# user-facing AI search; they aggregate content for training without
# offering retrieval-time citation. Any of these blocks can be
# removed if the studio later decides otherwise.

User-agent: CCBot
Disallow: /

User-agent: Omgilibot
Disallow: /

User-agent: ImagesiftBot
Disallow: /

Sitemap: https://muntin.digital/sitemap.xml