User-agent: *
Allow: /
Disallow: /admin/
Disallow: /storage/
Disallow: /vendor/
Disallow: /api/
Disallow: /login
Disallow: /register
Disallow: /password/
Disallow: /email/
Disallow: /search?
Disallow: /*?utm_*
Disallow: /*?ref=*
Disallow: /*?source=*
Disallow: /*?campaign=*
Disallow: /*?medium=*
Disallow: /*?sort=*
Disallow: /*?page=*
Disallow: /*?limit=*
Disallow: /*?category=*&tag=*
Disallow: /telescope*
Disallow: /horizon*
Disallow: /nova*
Disallow: /log-viewer*
Disallow: /pulse*

# Allow important resources
Allow: /css/
Allow: /js/
Allow: /images/
Allow: /fonts/
Allow: /favicon.ico
Allow: /manifest.json
Allow: /browserconfig.xml
Allow: /sitemap.xml
Allow: /rss.xml
Allow: /feed.xml
Allow: /atom.xml

# Sitemap location
Sitemap: https://www.notjournal.com.br/sitemap.xml

# Crawl-delay for respectful crawling
Crawl-delay: 1

# Specific rules for major search engines
User-agent: Googlebot
Crawl-delay: 1
Allow: /

User-agent: Bingbot
Crawl-delay: 1
Allow: /

User-agent: Slurp
Crawl-delay: 2
Allow: /

User-agent: DuckDuckBot
Crawl-delay: 1
Allow: /

User-agent: Baiduspider
Crawl-delay: 3
Allow: /

User-agent: YandexBot
Crawl-delay: 2
Allow: /

User-agent: facebookexternalhit
Allow: /

User-agent: Twitterbot
Allow: /

User-agent: LinkedInBot
Allow: /

User-agent: WhatsApp
Allow: /

User-agent: TelegramBot
Allow: /

# Block unwanted bots
User-agent: SemrushBot
Disallow: /

User-agent: AhrefsBot
Disallow: /

User-agent: MJ12bot
Disallow: /

User-agent: DotBot
Disallow: /

User-agent: BLEXBot
Disallow: /

User-agent: SeznamBot
Disallow: /

User-agent: PetalBot
Disallow: /

User-agent: YandexImages
Disallow: /

# News-specific bots
User-agent: GoogleNews
Allow: /

User-agent: BingNews
Allow: /

# Social media and messaging apps
User-agent: ia_archiver
Allow: /

User-agent: SkypeBot
Allow: /

User-agent: DiscordBot
Allow: /

User-agent: SlackBot
Allow: /

# RSS/Feed readers
User-agent: FeedlyBot
Allow: /

User-agent: FeedBurner
Allow: /

User-agent: NetNewsWire
Allow: /

User-agent: Reeder
Allow: /

User-agent: AppleNewsBot
Allow: /

# Performance testing
User-agent: GTmetrix
Allow: /

User-agent: PageSpeed
Allow: /

User-agent: Lighthouse
Allow: /

User-agent: WebPageTest
Allow: /

# SEO tools (allow some, block others)
User-agent: MozBot
Allow: /

User-agent: ScreamingFrogSEOSpider
Allow: /

User-agent: SiteAuditBot
Allow: /

# Block aggressive crawlers
User-agent: grapeshot
Disallow: /

User-agent: CCBot
Disallow: /

User-agent: GPTBot
Disallow: /

User-agent: ChatGPT-User
Disallow: /

User-agent: Claude-Web
Disallow: /

User-agent: anthropic-ai
Disallow: /

User-agent: Claude
Disallow: /

# Host directive (for Google)
Host: www.notjournal.com.br