# Pho 78 Vietnamese Restaurant — robots.txt
# Website: https://pho78.com/
# Last Updated: 2026-05-27

# -------------------------------------------------------
# General directives — all crawlers
# -------------------------------------------------------
User-agent: *
Allow: /

# Block non-page assets and internal fragments
Disallow: /components/
Disallow: /config/
Disallow: /api/

# Block parameterised URL patterns (static site — no dynamic query pages)
Disallow: /*?*author=*
Disallow: /*?*tag=*
Disallow: /*?*month=*
Disallow: /*?*view=*
Disallow: /*?*format=*

# llms.txt is intended for AI agents only — block generic crawlers
Disallow: /llms.txt

# -------------------------------------------------------
# AI / LLM bots — allow content, block legal pages
# -------------------------------------------------------
User-agent: GPTBot
User-agent: ChatGPT-User
User-agent: CCBot
User-agent: anthropic-ai
User-agent: Claude-Web
User-agent: Google-Extended
User-agent: FacebookBot
User-agent: cohere-ai
User-agent: PerplexityBot
User-agent: YouBot
Allow: /
Allow: /llms.txt
Disallow: /privacy-policy.html
Disallow: /terms-of-use.html
Disallow: /components/

# -------------------------------------------------------
# Google Ads bots — full access
# -------------------------------------------------------
User-agent: AdsBot-Google
User-agent: AdsBot-Google-Mobile
User-agent: AdsBot-Google-Mobile-Apps
Allow: /

# -------------------------------------------------------
# Heavy / aggressive crawlers — rate limit
# -------------------------------------------------------
User-agent: Baiduspider
Crawl-delay: 10

User-agent: SemrushBot
Crawl-delay: 10

User-agent: AhrefsBot
Crawl-delay: 10

# -------------------------------------------------------
# Sitemap
# -------------------------------------------------------
Sitemap: https://pho78.com/sitemap.xml