Sources

September 19, 2026 ยท View on GitHub

Fetched 2026-09-19 via Firecrawl map/scrape, Tavily map/crawl/extract/search, Context7 (/websites/typesafe_ai, /llmstxt/typesafe_ai_llms_txt), Mintlify .md + llms.txt / llms-full.txt, and https://api.typesafe.ai/openapi.json.

Live docs win if these notes drift.

Sites

HostWhat
https://typesafe.aiMarketing (Framer): manifesto, team, blog, legal
https://docs.typesafe.aiMintlify docs (109 sitemap URLs; JS SDK class pages omitted here)
https://console.typesafe.aiKeys + playground (login)
https://api.typesafe.aiPOST /v1/systemone, GET /v1/models, OpenAPI 0.2.0
https://evals.typesafe.aiPublic workflow evals
https://trust.typesafe.aiTrust center
https://status.typesafe.aiStatus

Concept pages ingested

  • https://docs.typesafe.ai/llms.txt
  • https://docs.typesafe.ai/llms-full.txt
  • introduction, quickstart, machine-learning-primer
  • concepts: system-one, state, how-to-build-with-system-one, use-case-map
  • primitives, choice, score, noul, advanced
  • confidence, api, models, model-jaggedness/jev-1.13
  • patterns + fan-out, confidence-routing, composite-scoring, intent-routing
  • agent-skill, migrating-to-v1, sdk (python/js overviews)
  • cookbooks listed in patterns.md (index + summaries; not every notebook cell)

Not copied: generated JS SDK class/interface pages, legal boilerplate.

Context7 libraries

  • /websites/typesafe_ai (primary)
  • /llmstxt/typesafe_ai_llms_txt
  • Nearby, unused: /websites/typesafe_ai_sdk_javascript, /saibimajdi/typesafe-dotnet-sdk