# robots.txt optimizado - mcmaqueda.com
# Europa & América | Actualizado julio 2026

# -----------------------------
# 1. BUSCADORES PRINCIPALES
# -----------------------------
User-agent: Googlebot
User-agent: Googlebot-Image
User-agent: AdsBot-Google
User-agent: Mediapartners-Google
User-agent: bingbot
User-agent: Applebot
Allow: /

# -----------------------------
# 2. REDES SOCIALES
# -----------------------------
User-agent: facebookexternalhit
User-agent: Pinterestbot
Allow: /

# -----------------------------
# 3. BOTS DE ASIA (bloquear)
# -----------------------------
User-agent: Baiduspider
User-agent: PetalBot
User-agent: Sogou web spider
User-agent: Bytespider
User-agent: YisouSpider
Disallow: /

# -----------------------------
# 4. BOT SOSPECHOSO "crawl"
# -----------------------------
User-agent: crawl
Disallow: /

# -----------------------------
# 5. ENTRENAMIENTO DE IA
# -----------------------------
User-agent: GPTBot
User-agent: CCBot
User-agent: Google-Extended
User-agent: anthropic-ai
User-agent: ClaudeBot
Disallow: /

# -----------------------------
# 6. CRAWLERS DE SEO (ralentizar)
# -----------------------------
User-agent: AhrefsBot
User-agent: SemrushBot
User-agent: SemrushBot-SI
User-agent: DotBot
User-agent: MJ12bot
User-agent: Barkrowler
Crawl-delay: 10
Allow: /

# -----------------------------
# 7. YANDEX (permitir con límite)
# -----------------------------
User-agent: YandexBot
User-agent: YandexImages
Crawl-delay: 5
Allow: /

# -----------------------------
# 8. REGLAS GENERALES DE WORDPRESS
# -----------------------------
User-agent: *
# Permitir recursos estáticos importantes
Allow: /wp-content/uploads/
Allow: /*.css$
Allow: /*.js$

# Bloquear zonas internas de WordPress
Disallow: /wp-admin/
Disallow: /wp-includes/
Disallow: /wp-content/plugins/
Disallow: /wp-content/themes/
Disallow: /cgi-bin/
Disallow: /xmlrpc.php
Disallow: /trackback
Disallow: /*/trackback
Disallow: /comments/
Disallow: /comment-page-
Disallow: /feed/
Disallow: /*/feed/
Disallow: /search
Disallow: /?s=
Disallow: /*?s=
Disallow: /page/
Disallow: /*/attachment/
Disallow: /?attachment_id=

# Ralentizar el resto de bots
Crawl-delay: 5

# -----------------------------
# 9. SITEMAPS
# -----------------------------
Sitemap: https://mcmaqueda.com/sitemap_index.xml
Sitemap: https://mcmaqueda.com/post-sitemap.xml
Sitemap: https://mcmaqueda.com/page-sitemap.xml
Sitemap: https://mcmaqueda.com/product-sitemap.xml
Sitemap: https://mcmaqueda.com/category-sitemap.xml
Sitemap: https://mcmaqueda.com/product_cat-sitemap.xml
Sitemap: https://mcmaqueda.com/product_tag-sitemap.xml

# LLMs
Allow: /llms.txt