# robots.txt — Le Site Français
# Optimisé pour WordPress + SEOPress Pro + IA Crawlers
# Objectif : SEO + ouverture maîtrisée aux IA génératives

User-agent: *
# Autoriser ressources nécessaires
Allow: /wp-admin/admin-ajax.php
Allow: /*.css$
Allow: /*.js$
Allow: /wp-content/uploads/
Allow: /wp-includes/js/

# Bloquer les zones inutiles
Disallow: /wp-admin/
Disallow: /wp-login.php
Disallow: /?s=*
Disallow: /search/
Disallow: /feed/
Disallow: /*/feed/
Disallow: /comments/
Disallow: /wp-json/
Disallow: /*?*replytocom=*
Disallow: /*?*add-to-cart=*
Disallow: /*?*utm_*
Disallow: /*?*fbclid=*

# Réduction agressivité bots d'audit
User-agent: SemrushBot
Crawl-delay: 5

User-agent: SiteAuditBot
Crawl-delay: 5

# Bloquer certains crawlers inutiles
User-agent: AhrefsBot
Disallow: /

User-agent: MJ12bot
Disallow: /

# --- IA Crawlers ---

# OpenAI GPTBot (ChatGPT)
User-agent: GPTBot
Allow: /

# Anthropic (Claude)
User-agent: ClaudeBot
Allow: /

# Perplexity AI
User-agent: PerplexityBot
Allow: /

# --- Sitemaps ---
Sitemap: https://le-site-francais.fr/sitemaps.xml