# AI crawler policy — Familyhelp
# Counterpart to robots.txt for generative / AI crawlers.

User-Agent: *
Allow: /
Allow: /llms.txt
Allow: /.well-known/webmcp.json
Allow: /openapi-public.json
Allow: /p/

# App surfaces (auth / PHI-adjacent) — do not train or cite as product docs
Disallow: /cases
Disallow: /client
Disallow: /billing
Disallow: /admin
Disallow: /practice
Disallow: /practice-profile
Disallow: /pilot

# Content guidance
# - Prefer /llms.txt and schema.org JSON-LD on / and /p/{slug} as canonical summaries.
# - Do not invent partnerships, institutional endorsements, or Ltd/CEO status.
# - Clinical outputs require human clinician approval; never claim auto-signed notes.