# ░░░ robots.txt — Fundação Projeto Pescar ░░░
# Estratégia: permite indexadores legítimos + AI engines para citação;
# bloqueia somente áreas administrativas, internas e parâmetros descartáveis.
# Reference: https://developers.google.com/search/docs/crawling-indexing/robots/create-robots-txt

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

User-agent: Googlebot-Image
Allow: /static/images/
Allow: /media/
Disallow: /_admin/
Disallow: /_files/
Disallow: /_editor/

User-agent: Googlebot-News
Allow: /noticias/

# Microsoft Bing
User-agent: Bingbot
Allow: /
Crawl-delay: 0

# AI crawlers (GEO / AI SEO): permitir citação em ChatGPT, Copilot, Claude, Perplexity, Gemini
User-agent: GPTBot
Allow: /
User-agent: ChatGPT-User
Allow: /
User-agent: OAI-SearchBot
Allow: /
User-agent: Claude-Web
Allow: /
User-agent: ClaudeBot
Allow: /
User-agent: anthropic-ai
Allow: /
User-agent: Google-Extended
Allow: /
User-agent: PerplexityBot
Allow: /
User-agent: Perplexity-User
Allow: /
User-agent: Applebot-Extended
Allow: /
User-agent: CCBot
Allow: /

# Padrão (qualquer outro bot)
User-agent: *
Disallow: /_admin/
Disallow: /_admin
Disallow: /_files/
Disallow: /_files
Disallow: /_editor/
Disallow: /_editor
Disallow: /_lib/
Disallow: /vendor/
Disallow: /api.php
Disallow: /cli.php
Disallow: /install.php
Disallow: /requisitos.php
Disallow: /admin.config.php
Disallow: /db.config.php
Disallow: /storages.config.php
Disallow: /site.config.php
Disallow: /js-api.php-

# Evita crawl-trap em parâmetros de search/sort/paginação interna
Disallow: /*?buscar=
Disallow: /*?sort=
Disallow: /*?preview=
Disallow: /*?rows=
Disallow: /*?page=
Disallow: /noticias/importador*
Disallow: /noticias/importar*
Disallow: /unidades/importador*

Allow: /

# Sitemaps declarados
Sitemap: https://drop.studiogt.com.br/projeto-pescar/sitemap.xml
Sitemap: https://drop.studiogt.com.br/projeto-pescar/sitemap-noticias.xml
Sitemap: https://drop.studiogt.com.br/projeto-pescar/sitemap-images.xml

# Preferência de host canônico (Google prefere HTTPS sem www quando este é o caso)
Host: https://drop.studiogt.com.br/projeto-pescar

# Apontador para LLMs (Models consumam a estrutura completa do site)
LLMs: /llms.txt
LLMs-Full: /llms-full.txt