# The Durability Curve — durabilitycurve.com
#
# Policy (set 2026-08-23): OPEN TO AI. This site may be indexed for search,
# used in real-time AI answers (RAG / grounding), and used to train or
# fine-tune models. Every well-behaved crawler is allowed.
#
# Cloudflare's managed robots.txt is DISABLED (dashboard → the domain →
# Manage AI bot access → "Manage your robots.txt" → Disable), so THIS file is
# authoritative: what you read here is what serves. Edit here; it deploys with
# the site. (This reverses the earlier "training refused" policy — the content
# is already public on Substack, and the goal is reach and citation, not IP
# reservation. The paid instruments stay gated by X-Robots-Tag, see below.)
#
# The Content-Signal line is the machine-readable statement of the policy
# (Content Signals Policy, contentsignals.org):
#   search   = yes  — build a search index; return links and short excerpts
#   ai-input = yes  — use the content to ground real-time AI answers
#   ai-train = yes  — train or fine-tune models on the content

User-agent: *
Content-Signal: search=yes,ai-input=yes,ai-train=yes
Allow: /

# NOTE — the paid instruments are deliberately NOT named here.
# This file is world-readable, so a `Disallow` on a paid tool publishes the exact
# unguessable URL that IS the gate. It is also self-defeating: a crawler blocked here
# never fetches the page, so it never reads the noindex, and a URL discovered any other
# way can still surface as a bare URL result. The gate is enforced where it works —
# `X-Robots-Tag: noindex, nofollow` in public/_headers (verified live 2026-07-26), plus a
# noindex meta in the instrument's own head. Do not add paid paths back to this file.

# The /ghost/ archive was retired 2026-10-07 (S431): its URLs 301 to main-site pages
# (public/_redirects). Do not re-add it or its sitemap.
Sitemap: https://durabilitycurve.com/sitemap-index.xml