User-agent: *

Disallow: /sfCounter/*
Disallow: /sfRating/*
Disallow: /sfRecommend/*
Disallow: /sfComment/score/id/*
Disallow: /sfReadLater/*

# IAs
# Google (Bard / Gemini). 
# Permitido experimentalmente
User-agent: Google-Extended
Disallow:

#Facebook agent (NoIA)
User-agent: facebookexternalhit 
Allow: /

# Amazon (IA relacionada con Alexa y modelos)
User-agent: AmazonBot
Disallow: /

# Permitimos el bot publicitario de amazon
User-agent: AmazonAdBot
Disallow:

#Recomendacion https://knownagents.com/agents/
#AI Data Scrapers (completa)
#Sección Undocumented AI Agents (parcial)
# AI data scrappers for training LLMs
User-agent: anthropic-ai
Disallow: /
User-agent: AI2Bot
Disallow: /
User-agent: Ai2Bot-Dolma
Disallow: /
User-agent: ApifyBot
Disallow: /
User-agent: ApifyWebsiteContentCrawler
Disallow: /
User-agent: Applebot-Extended
Disallow: /
User-agent: Bytespider
Disallow: /
User-agent: CCBot
Disallow: /
User-agent: ChatGLM-Spider
Disallow: /
User-agent: ClaudeBot
Disallow: /
User-agent: cohere-ai
Disallow: /
User-agent: cohere-training-data-crawler
Disallow: /
User-agent: Cotoyogi
Disallow: /
User-agent: Crawl4AI
Disallow: /
User-agent: Datenbank Crawler
Disallow: /
User-agent: DeepSeekBot
Disallow: /
User-agent: Diffbot
Disallow: /
User-agent: FacebookBot
Allow: /
User-agent: FirecrawlAgent
Disallow: /
User-agent: GPTBot
Disallow: /
User-agent: iAskBot
Disallow: /
User-agent: iaskspider
Disallow: /
User-agent: ICC-Crawler
Disallow: /
User-agent: Kangaroo Bot
Disallow: /
User-agent: KunatoCrawler
Disallow: /
User-agent: laion-huggingface-processor
Disallow: /
User-agent: LCC
Disallow: /
User-agent: meta-externalagent
Disallow: /
User-agent: netEstate Imprint Crawler
Disallow: /
User-agent: omgili
Disallow: /
User-agent: omgilibot
Disallow: /
User-agent: PanguBot
Disallow: /
User-agent: SBIntuitionsBot
Disallow: /
User-agent: SemrushBot-OCOB
Disallow: /
User-agent: Spider
Disallow: /
User-agent: TavilyBot
Disallow: /
User-agent: Timpibot
Disallow: /
User-agent: VelenPublicWebCrawler
Disallow: /
User-agent: webzio-extended
Disallow: /
User-agent: WRTNBot
Disallow: /

# Yahoo's Slurp Robot - Please wait 15 seconds in between visits 
User-agent: slurp 
Crawl-delay: 15

# MSN Robot - Please wait 15 seconds in between visits 
User-agent: msnbot 
Crawl-delay: 15

#Generales
Sitemap: https://www.dbalears.cat/sitemap.xml
Sitemap: https://www.dbalears.cat/googlenews.xml
Sitemap: https://www.dbalears.cat/image-sitemap.xml
#Autores
Sitemap: https://www.dbalears.cat/autors.xml
#Hemeroteca
Sitemap: https://www.dbalears.cat/noticies/sitemapIndex.xml
#Elecciones:
Sitemap: https://www.dbalears.cat/elecciones.xml
#Media
Sitemap: https://www.dbalears.cat/fotogalerias.xml
Sitemap: https://www.dbalears.cat/videos.xml