# ==========================================
# stavebne-dopyty.sk - Robots.txt
# Anti-scraping + SEO + GEO optimized
# ==========================================

User-agent: *
Allow: /
Crawl-delay: 2

# Sitemap & AI files
Sitemap: https://stavebne-dopyty.sk/api/local-seo/sitemap/stavebne-dopyty/index.xml
Sitemap: https://stavebne-dopyty.sk/sitemap.xml

# Disallow admin and private pages
Disallow: /admin
Disallow: /profile
Disallow: /onboarding
Disallow: /chat
Disallow: /auth/
Disallow: /dashboard

# Disallow API endpoints (private/transactional)
# Allow listed FIRST — specific rules win over generic `/api/` in Googlebot RFC.
Allow: /api/seo/
Allow: /api/local-seo/sitemap/
Allow: /api/local-seo/robots/
Allow: /api/uploads/
Allow: /api/llms.txt
Allow: /api/llms-full.txt
Disallow: /api/
Disallow: /api/demands
Disallow: /api/offers
Disallow: /api/users
Disallow: /api/companies
Disallow: /api/admin

# Allow public SEO assets
Allow: /static/
Allow: /*.js$
Allow: /*.css$
Allow: /*.png$
Allow: /*.jpg$
Allow: /*.webp$
Allow: /sitemap.xml
Allow: /llms.txt
Allow: /llms-full.txt

# Allow SEO landing pages for indexing
Allow: /kategoria/
Allow: /lokalita/
Allow: /ako-to-funguje
Allow: /faq
Allow: /blog
Allow: /pre-zakaznikov
Allow: /pre-poskytovatelov
Allow: /cennik
Allow: /stavebny-dennik
Allow: /fakturacia
Allow: /sklad
Allow: /dopyty

# ==========================================
# Block known scraping bots and tools
# ==========================================

User-agent: MJ12bot
Disallow: /

User-agent: AhrefsBot
Disallow: /

User-agent: SemrushBot
Disallow: /

User-agent: DotBot
Disallow: /

User-agent: BLEXBot
Disallow: /

User-agent: MegaIndex.ru
Disallow: /

User-agent: Bytespider
Disallow: /

User-agent: PetalBot
Disallow: /

User-agent: SeznamBot
Crawl-delay: 5

User-agent: YandexBot
Crawl-delay: 5

User-agent: DataForSeoBot
Disallow: /

User-agent: Scrapy
Disallow: /

User-agent: python-requests
Disallow: /

User-agent: curl
Disallow: /

User-agent: wget
Disallow: /

User-agent: httrack
Disallow: /

User-agent: harvest
Disallow: /

User-agent: WebCopier
Disallow: /

User-agent: Offline Explorer
Disallow: /

User-agent: Teleport
Disallow: /

User-agent: emailcollector
Disallow: /

User-agent: webbandit
Disallow: /

User-agent: ia_archiver
Disallow: /

User-agent: Nutch
Disallow: /

# ==========================================
# AI Search Engines - FULL ACCESS (GEO)
# ==========================================

User-agent: GPTBot
Allow: /
Allow: /kategoria/
Allow: /lokalita/
Allow: /ako-to-funguje
Allow: /faq
Allow: /blog
Allow: /llms.txt
Allow: /llms-full.txt
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin
Disallow: /profile
Crawl-delay: 2

User-agent: OAI-SearchBot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin
Crawl-delay: 2

User-agent: ChatGPT-User
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin

User-agent: Google-Extended
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin

User-agent: PerplexityBot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin
Crawl-delay: 2

User-agent: ClaudeBot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin
Crawl-delay: 2

User-agent: anthropic-ai
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/

User-agent: CCBot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin
Crawl-delay: 3

User-agent: FacebookBot
Allow: /

User-agent: Bingbot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Crawl-delay: 2

User-agent: Googlebot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin
Disallow: /profile

User-agent: Googlebot-Image
Allow: /
Allow: /api/uploads/
Disallow: /api/
Disallow: /admin

User-agent: Applebot
Allow: /
Allow: /api/seo/
Allow: /api/uploads/
Disallow: /api/