# ============================================================
# SWJARVIS.COM — robots.txt
# © 2026 SWJARVIS S.R.L. — P.IVA 02308560503
# ============================================================
# Rif: Robots Exclusion Protocol — https://www.rfc-editor.org/rfc/rfc9309
# Rif: Google — https://developers.google.com/search/docs/crawling-indexing/robots/intro
# Rif: Bing — https://www.bing.com/webmasters/help/how-to-create-a-robots-txt-file
# ============================================================

# --- Motori di ricerca tradizionali ---
User-agent: Googlebot
Allow: /
Crawl-delay: 1

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

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

User-agent: DuckDuckBot
Allow: /

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

# --- AI Crawler / LLM Agents ---
# Rif: https://platform.openai.com/docs/bots
# Rif: https://docs.anthropic.com/en/docs/about-claude/web-crawler
# Rif: https://developers.google.com/search/docs/crawling-indexing/google-extended
# Rif: https://docs.perplexity.ai/docs/perplexitybot

User-agent: GPTBot
Allow: /

User-agent: ChatGPT-User
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: Applebot-Extended
Allow: /

User-agent: cohere-ai
Allow: /

User-agent: meta-externalagent
Allow: /

User-agent: Bytespider
Allow: /

User-agent: CCBot
Allow: /

# --- Tutti gli altri ---
User-agent: *
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /wp-admin/
Disallow: /wp-login.php
Disallow: /*.bak$
Disallow: /*.sql$
Disallow: /*.log$

# --- Sitemap ---
Sitemap: https://swjarvis.com/sitemap.xml
Sitemap: https://swjarvis.com/sitemap-ai.xml

# --- File AI/LLM ---
# llms.txt: https://swjarvis.com/llms.txt
# llms-full.txt: https://swjarvis.com/llms-full.txt
# ai.txt: https://swjarvis.com/ai.txt
# knowledge.jsonld: https://swjarvis.com/knowledge.jsonld