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

# Block parameter variants of signin/signup (prevents index bloat from
# ?redirect= / ?source= variants). Bare URLs remain indexable.
Disallow: /auth/signin?
Disallow: /auth/signup?

# OAuth callbacks
Disallow: /auth/signin/

# Transactional auth flows
Disallow: /auth/verify-email
Disallow: /auth/forgot-password
Disallow: /auth/set-password
Disallow: /auth/profile-setup
Disallow: /auth/app-download

# Dashboard (post-login)
Disallow: /partners/

# AI crawlers — explicit allow (we want to be citable in AI search).
# Mix of training crawlers (GPTBot, ClaudeBot, etc.) and user-initiated
# browse/search bots (ChatGPT-User, Claude-Web, Perplexity-User).
# Stacked User-agent lines form a single group sharing the rules below
# (pattern modeled on Cloudflare's robots.txt).

# OpenAI (training + browsing + search citations)
User-agent: GPTBot
User-agent: ChatGPT-User
User-agent: OAI-SearchBot

# Anthropic (training + browsing)
User-agent: ClaudeBot
User-agent: anthropic-ai
User-agent: Claude-Web

# Google AI (training opt-in; core Googlebot still uses the * group)
User-agent: Google-Extended

# Apple Intelligence (training opt-in; core Applebot still uses the * group)
User-agent: Applebot-Extended

# Perplexity (indexing + user-initiated browse)
User-agent: PerplexityBot
User-agent: Perplexity-User

# Meta AI (Llama training, AI features in WhatsApp/Instagram/Facebook)
User-agent: Meta-ExternalAgent
User-agent: FacebookBot

# Other major AI crawlers
User-agent: Amazonbot
User-agent: Bytespider
User-agent: cohere-ai
User-agent: CCBot
User-agent: YouBot
Allow: /

Sitemap: https://www.impakt.com/sitemap.xml