# robots.txt – SEOX / Publisher - 2026-04

# ------------------------------------------------------------------
# BLOQUEIO GERAL & ROTAS ADMINISTRATIVAS
# ------------------------------------------------------------------
User-agent: *
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php
Disallow: /wp-login.php
Disallow: /xmlrpc.php
Disallow: /?s=
Disallow: /*?s=
Disallow: /search
Disallow: /search$
Disallow: /search/
Disallow: /trackback/
Disallow: /readme.html
Disallow: /license.txt
Disallow: /cdn-cgi/
Disallow: /private/
Disallow: /feed/
Disallow: /*/feed/

# ------------------------------------------------------------------
# CRAWLERS SEO (herdam tudo acima + restrição extra de /wp-json/)
# Regras replicadas porque robots.txt NÃO faz merge entre grupos
# (RFC 9309 + Google docs)
# ------------------------------------------------------------------
User-agent: Googlebot
User-agent: Bingbot
User-agent: DuckDuckBot
User-agent: Slurp
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php
Disallow: /wp-login.php
Disallow: /xmlrpc.php
Disallow: /?s=
Disallow: /*?s=
Disallow: /search
Disallow: /search$
Disallow: /search/
Disallow: /trackback/
Disallow: /readme.html
Disallow: /license.txt
Disallow: /cdn-cgi/
Disallow: /private/
Disallow: /feed/
Disallow: /*/feed/
Disallow: /wp-json/

# ------------------------------------------------------------------
# FACEBOOK/WHATSAPP oEmbed preview — fail-safe explícito
# ------------------------------------------------------------------
User-agent: facebookexternalhit
Allow: /wp-json/oembed/
Disallow:

# ------------------------------------------------------------------
# AI CRAWLERS & LLM SCRAPERS
# ------------------------------------------------------------------
User-agent: CCBot
User-agent: PerplexityBot
User-agent: anthropic-ai
User-agent: ClaudeBot
User-agent: Omgilibot
User-agent: Diffbot
User-agent: Applebot-Extended
User-agent: GPTBot
User-agent: ChatGPT-User
User-agent: OAI-SearchBot
User-agent: Google-Extended
User-agent: meta-externalagent
User-agent: LCC
Disallow: /

# ------------------------------------------------------------------
# SEO / SCRAPING BOTS AGRESSIVOS
# ------------------------------------------------------------------
User-agent: Amazonbot
User-agent: Amzn-SearchBot
User-agent: PetalBot
User-agent: SemrushBot
User-agent: SemrushBot-SA
User-agent: Bytespider
User-agent: SeobilityBot
User-agent: AhrefsBot
User-agent: BLEXBot
User-agent: DotBot
User-agent: linkdexbot
User-agent: Scrapy
User-agent: YandexBot
User-agent: Baiduspider
User-agent: Sogou
User-agent: MJ12bot
User-agent: DataForSeoBot
User-agent: ImageSiftBot
User-agent: MegaIndex
User-agent: MauiBot
Disallow: /

# ------------------------------------------------------------------
# SITEMAPS
# ------------------------------------------------------------------
Sitemap: https://www.giromarilia.com.br/wp-content/sitemaps/sitemap_index.xml
Sitemap: https://www.giromarilia.com.br/wp-content/sitemaps/news-sitemap.xml