# robots.txt pour WordPress
# Mise à jour : Novembre 2025
# Site : https://carnetdemarseille.com

# === ROBOTS AUTORISÉS (Moteurs de recherche principaux) ===

User-agent: Googlebot
User-agent: bingbot
User-agent: YandexBot
User-agent: DuckDuckBot
User-agent: Applebot
User-agent: facebookexternalhit
User-agent: Pinterestbot
User-agent: archive.org_bot

# Tout autoriser pour ces robots
Allow: /

# Sauf les dossiers sensibles WordPress
Disallow: /wp-admin/
Disallow: /wp-includes/
Disallow: /wp-content/plugins/
Disallow: /wp-content/cache/
Disallow: /wp-content/themes/*/includes/

# MAIS autoriser les ressources publiques essentielles
Allow: /wp-admin/admin-ajax.php
Allow: /wp-content/uploads/
Allow: /wp-content/themes/*/css/
Allow: /wp-content/themes/*/js/
Allow: /wp-content/themes/*/images/
Allow: /wp-content/themes/*/assets/
Allow: /wp-includes/css/
Allow: /wp-includes/js/

# Bloquer les pages de recherche et paramètres inutiles
Disallow: /?s=
Disallow: /search/
Disallow: *?replytocom=
Disallow: /xmlrpc.php
Disallow: /feed/
Disallow: /trackback/
Disallow: /comments/feed/
Disallow: /author/
Disallow: /tag/
Disallow: *?attachment_id=

# === ROBOTS SEO (avec limitations) ===

User-agent: AhrefsBot
User-agent: SemrushBot
User-agent: MJ12bot
User-agent: Majestic12
User-agent: DotBot
Crawl-delay: 10
Allow: /
Disallow: /wp-admin/
Disallow: /wp-includes/
Disallow: /wp-content/plugins/
Disallow: /wp-content/cache/
Disallow: /wp-content/themes/
Disallow: /?s=
Disallow: /feed/

# === TOUS LES AUTRES ROBOTS BLOQUÉS ===

User-agent: *
Disallow: /

# === SITEMAP ===

Sitemap: https://carnetdemarseille.com/sitemap.xml
Sitemap: https://carnetdemarseille.com/sitemap_index.xml