# robots.txt — Sankhya Farms
# https://sankhyafarms.com

# ── All crawlers (default) ──────────────────────────────────────────
User-agent: *
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/
Disallow: /*.json$

Sitemap: https://sankhyafarms.com/sitemap.xml

# Optional machine-readable context file for compatible AI systems.
# Google Search does not use llms.txt as a ranking or AI-visibility signal.
# https://sankhyafarms.com/llms.txt

# ── Google ───────────────────────────────────────────────────────────
User-agent: Googlebot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/
Disallow: /*.json$

User-agent: Google-Extended
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/
Disallow: /*.json$

User-agent: Googlebot-Image
Allow: /
Disallow: /api/

# ── Microsoft / Bing ─────────────────────────────────────────────────
User-agent: Bingbot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Apple ────────────────────────────────────────────────────────────
User-agent: Applebot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: Applebot-Extended
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── OpenAI ───────────────────────────────────────────────────────────
# GPTBot controls potential model-training crawl; OAI-SearchBot controls ChatGPT Search discovery.
User-agent: GPTBot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: OAI-SearchBot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: ChatGPT-User
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/


# ── Meta ──────────────────────────────────────────────────────────────
User-agent: Meta-ExternalAgent
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: Meta-ExternalFetcher
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Cohere ────────────────────────────────────────────────────────────
User-agent: cohere-ai
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Amazon / Alexa ───────────────────────────────────────────────────
User-agent: Amazonbot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── ByteDance / TikTok ───────────────────────────────────────────────
User-agent: Bytespider
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── AI2 (Allen Institute) ────────────────────────────────────────────
User-agent: AI2Bot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Diffbot ───────────────────────────────────────────────────────────
User-agent: Diffbot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Common Crawl (training corpus for GPT, Llama, Mistral, most open models) ─
User-agent: CCBot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Anthropic ─────────────────────────────────────────────────────────
User-agent: anthropic-ai
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: ClaudeBot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: Claude-SearchBot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

User-agent: Claude-Web
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/

# ── Perplexity ────────────────────────────────────────────────────────
User-agent: PerplexityBot
Allow: /
Disallow: /api/
Disallow: /admin/
Disallow: /live/
Disallow: /login/