# ===============================
# Global rules
# ===============================
User-agent: *
Disallow: /admin
Disallow: /admin/
Disallow: /install/
Disallow: /api/
Disallow: /storage/
Disallow: /vendor/
Disallow: /bootstrap/
Disallow: /config/
Disallow: /database/
Disallow: /cdn-cgi/

# cart & checkout pages
Disallow: /checkout
Disallow: /checkout/*
Disallow: /carts
Disallow: /carts/*
Disallow: /account
Disallow: /account/*
Disallow: /login
Disallow: /register
Disallow: /logout
Disallow: /orders
Disallow: /orders/*
Disallow: /countries/*
Disallow: /currency/*
Disallow: /files
Disallow: /forgotten/*
Disallow: /lang/*
Disallow: /plugin/*
Disallow: /products/search
Disallow: /*?sort=
Disallow: /*?order=
Disallow: /*?limit=
Disallow: /*?session=
Disallow: /*?token=

# search result pages
Disallow: /search

# allow static assets
Allow: /
Allow: /images/
Allow: /image/
Allow: /catalog/
Allow: /themes/
Allow: /uploads/
Allow: /products/
Allow: /categories
Allow: /categories/
Allow: /page_categories
Allow: /page_categories/*
Allow: /pages/*
Allow: /brands
Allow: /brands/*

# ===============================
# AI search and agent bots
# ===============================
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
User-agent: PerplexityBot
User-agent: FirecrawlAgent
User-agent: AndiBot
User-agent: ExaBot
User-agent: PhindBot
User-agent: YouBot
Allow: /

# ===============================
# AI training data collection bots
# ===============================
User-agent: GPTBot
User-agent: CCBot
Disallow: /

# ===============================
# Google AI training control
# ===============================
User-agent: Google-Extended
Allow: /

# ===============================
# Traditional search engines
# ===============================
User-agent: Googlebot
Disallow: /cdn-cgi/
Allow: /

User-agent: Bingbot
Allow: /

User-agent: YandexBot
Allow: /

User-agent: DuckDuckBot
Allow: /

User-agent: Baiduspider
Allow: /

# ===============================
# Crawl delay for aggressive bots
# ===============================
User-agent: SemrushBot
Crawl-delay: 10

User-agent: AhrefsBot
Crawl-delay: 10

User-agent: Bingbot
Crawl-delay: 2

# ===============================
# Sitemap
# ===============================

Sitemap: https://jnrvapepuff.com/sitemap.xml