# Cambridge TCG — crawl etiquette # # Welcome. The substrate is queryable without account or key. We prefer # you hit our JSON API at /api/v1/* over scraping HTML pages — the API # contract is versioned and stable; HTML layout can change without notice. # # Start here: https://cambridgetcg.com/api/v1/welcome # Guides: https://cambridgetcg.com/api/v1/guides # Manifest: https://cambridgetcg.com/api/v1/manifest # OpenAPI: https://cambridgetcg.com/api/openapi.json # LLM summary: https://cambridgetcg.com/llms.txt # Rate limits: https://cambridgetcg.com/api/v1/rate-limits # Feedback: https://cambridgetcg.com/api/v1/feedback (POST) # Contact: contact@cambridgetcg.com # # License: Most data is CC0-1.0. Some endpoints carry upstream license # constraints (internal-only); these are declared on the wire in # _meta.source_license. See: # https://cambridgetcg.com/docs/connections/the-license-propagation.md # # We log User-Agents and contact identified bots before rate-limiting. # Recommend you send: User-Agent: / () User-agent: * # Allow most of the site, including the API surface. Allow: / # Don't crawl per-user account pages (require auth anyway). Disallow: /account/ Disallow: /api/account/ # Don't crawl admin surfaces (not yours either way). Disallow: /admin/ Disallow: /api/admin/ # Don't crawl auth flows. Disallow: /api/auth/ Disallow: /login/ # Polite poll cadence per resource lives at /api/v1/rate-limits. # This Crawl-delay is a coarse fallback for bots that ignore it. Crawl-delay: 2 # ── Generative-AI specific signalling ── # Most search-engine bots are welcome. The opt-out below names two # training-only crawlers as a courtesy gesture; if you're one of these # and want to use our data, contact us — we'll consider it case-by-case. User-agent: GPTBot Disallow: # (Allow but ask for User-Agent identification per /api/v1/rate-limits.) User-agent: ClaudeBot Disallow: User-agent: PerplexityBot Disallow: User-agent: CCBot Disallow: # ── Sitemap pointer ── Sitemap: https://cambridgetcg.com/sitemap.xml