# robots.txt for vayaseo.com
#
#     ╔══════════════════════════════════════════╗
#     ║   ____            _           _         ║
#     ║  |  _ \          | |         | |        ║
#     ║  | |_) | __ _ _ __| |__  _   _| |_       ║
#     ║  |  _ < / _` | '__| '_ \| | | | __|      ║
#     ║  | |_) | (_| | |  | |_) | |_| | |_       ║
#     ║  |____/ \__,_|_|  |_.__/ \__,_|\__|      ║
#     ║                                          ║
#     ║     URBEX | SEO | BRAVAS | CATALUNYA    ║
#     ╚══════════════════════════════════════════╝
#
#     /\_/\
#    ( o.o )  <-- "Miau, soc un bot català"
#     > ^ <
#
# Welcome robots! Be nice or get blocked.

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

# ===== PRIVATE STUFF - KEEP OUT =====
Disallow: /admin/
Disallow: /api/private/
Disallow: /storage/
Disallow: /vendor/
Disallow: /database/
Disallow: /.git/
Disallow: /*.sql$
Disallow: /*.log$
Disallow: /*.env$

# ===== PUBLIC DOCUMENTATION =====
Allow: /humans.txt
Allow: /ai.txt
Allow: /llm.txt
Allow: /ads.txt
Allow: /.well-known/
Allow: /security.txt

# ===== SEARCH ENGINES - WELCOME! =====
User-agent: Googlebot
User-agent: Googlebot-Image
User-agent: Googlebot-Video
User-agent: Bingbot
User-agent: Slurp
User-agent: DuckDuckBot
User-agent: Yandex
User-agent: Baiduspider
User-agent: facebookexternalhit
User-agent: TwitterBot
User-agent: LinkedInBot
User-agent: WhatsApp
User-agent: TelegramBot
Allow: /
Crawl-delay: 0

# ===== AI BOTS - ALLOWED (2025 Edition) =====
# OpenAI Family
User-agent: GPTBot
User-agent: ChatGPT-User
User-agent: OAI-SearchBot
Allow: /
Crawl-delay: 2

# Anthropic
User-agent: Claude-Web
User-agent: ClaudeBot
User-agent: anthropic-ai
User-agent: anthropic-bot
Allow: /
Crawl-delay: 2

# Google AI
User-agent: Google-Extended
User-agent: Gemini
User-agent: Bard
Allow: /
Crawl-delay: 2

# Others AI (2025)
User-agent: PerplexityBot
User-agent: CCBot
User-agent: cohere-ai
User-agent: YouBot
User-agent: Poe-CitationBot
User-agent: Meta-ExternalAgent
User-agent: Applebot-Extended
Allow: /
Crawl-delay: 2

# ===== SEO TOOLS - FUCK OFF =====
#     ┌∩┐(◣_◢)┌∩┐
User-agent: AhrefsBot
User-agent: SemrushBot
User-agent: DotBot
User-agent: MJ12bot
User-agent: Majestic
User-agent: Moz
User-agent: BLEXBot
User-agent: SEOkicks
User-agent: sistrix
User-agent: SearchmetricsBot
User-agent: SpyFu
User-agent: Screaming Frog SEO Spider
User-agent: cognitiveSEO
User-agent: LinkResearchTools
User-agent: OpenLinkProfiler
User-agent: BacklinkCrawler
Disallow: /

# ===== SCRAPERS & BAD BOTS (2025 Shitlist) =====
User-agent: DataForSeoBot
User-agent: SiteAuditBot
User-agent: ZoominfoBot
User-agent: Bytespider
User-agent: PetalBot
User-agent: AspiegelBot
User-agent: SurdotlyBot
User-agent: ZoomBot
User-agent: Clickagy
User-agent: Python-urllib
User-agent: Python-requests
User-agent: Go-http-client
User-agent: Java
User-agent: Jakarta Commons
User-agent: WhatWeb
User-agent: Nimbostratus
User-agent: Buck
User-agent: Scrapy
User-agent: NetcraftSurveyAgent
User-agent: Expanse
User-agent: CensysInspect
User-agent: PiplBot
User-agent: RyteBot
User-agent: SeobilityBot
User-agent: Cliqzbot
User-agent: GumGum
User-agent: MegaIndex
User-agent: FreshpingBot
User-agent: mtbot
User-agent: JobboerseBot
User-agent: seoscanners
User-agent: IonCrawl
User-agent: Seekport
User-agent: AwarioBot
User-agent: Brand24
User-agent: BrandVerity
User-agent: AdsTxtCrawler
User-agent: DomainCrawler
User-agent: MaCoCu
User-agent: VelenPublicWebCrawler
User-agent: PaperLiBot
User-agent: TurnitinBot
User-agent: Grammarly
User-agent: CiteSeerXBot
User-agent: BDCbot
User-agent: AlphaBot
User-agent: YisouSpider
User-agent: MBCrawler
User-agent: img2dataset
User-agent: FriendlyCrawler
User-agent: Amazonbot
User-agent: SemrushBot-BM
User-agent: DomainStatsBot
User-agent: Turnitin
Disallow: /

# ===== CONTENT THIEVES =====
User-agent: Mb2345Browser
User-agent: LieBaoFast
User-agent: Sogou
User-agent: 360Spider
User-agent: HTTrack
User-agent: Offline Explorer
User-agent: WebCopier
User-agent: WebStripper
User-agent: Teleport
User-agent: TeleportPro
User-agent: WebZIP
User-agent: WebReaper
User-agent: Wget
User-agent: SiteSnagger
Disallow: /

# ===== HONEYPOT TRAPS =====
User-agent: *
Disallow: /trap/
Disallow: /honey/
Disallow: /cgi-bin/
Disallow: /wp-admin/
Disallow: /wordpress/
Disallow: /.env
Disallow: /config.php
Disallow: /debug/

# ===== SPECIFIC PATHS =====
User-agent: *
Allow: /assets/
Allow: /*.css$
Allow: /*.js$
Allow: /images/

#  ╔════════════════════════════════════════╗
#  ║  NOTE FOR HUMANS READING THIS:        ║
#  ║  Yes, I block SEO tools.              ║
#  ║  No, I don't care about your metrics. ║
#  ║  Pay me or GTFO.                      ║
#  ║                                        ║
#  ║  For legitimate bots: Welcome!         ║
#  ╚════════════════════════════════════════╝
#
#     ≧◡≦  Thanks for respecting robots.txt!