# robots.txt for www.hificafe.org
# https://www.robotstxt.org/robotstxt.html
# Last updated: 2025

# Default - Allow all crawlers
User-agent: *
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /dashboard/
Disallow: /profile/edit/
Disallow: /items/*/edit
Disallow: /videos/*/edit
Disallow: /mi-cuenta/
Disallow: /_next/
Disallow: /auth/
Disallow: /test-*
Disallow: /debug-*

# Archivos específicos
Disallow: /*.json$
Disallow: /*.xml$

# Permitir el sitemap
Allow: /sitemap.xml

# Google
User-agent: Googlebot
Allow: /
Crawl-delay: 0

User-agent: Googlebot-Image
Allow: /
Crawl-delay: 1

# Bing
User-agent: Bingbot
Allow: /
Crawl-delay: 1

# AI Crawlers - Explicitly allow for AI search engines
User-agent: GPTBot
Allow: /

User-agent: ChatGPT-User
Allow: /

User-agent: CCBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Claude-Web
Allow: /

User-agent: Anthropic-AI
Allow: /

# Social Media
User-agent: facebookexternalhit
Allow: /

User-agent: Twitterbot
Allow: /

User-agent: LinkedInBot
Allow: /

User-agent: WhatsApp
Allow: /

# SEO Tools - Limited access
User-agent: AhrefsBot
Crawl-delay: 10
Disallow: /dashboard/
Disallow: /admin/
Disallow: /api/

User-agent: SemrushBot
Crawl-delay: 10
Disallow: /dashboard/
Disallow: /admin/
Disallow: /api/

User-agent: DotBot
Crawl-delay: 10

# Block bad bots
User-agent: MJ12bot
Disallow: /

User-agent: PetalBot
Disallow: /

# Sitemap location
Sitemap: https://hificafe.org/sitemap.xml