Building structured AI decision models
The documentation explains how to use TypeSafe AI's System One models with typed questions and state to generate structured, code-consumable decisions.
All sources
TypeSafe AI
https://docs.typesafe.aiTypeSafe AI Documentation for AI Agents & MCP
TypeSafe AI documentation guides AI agents in building applications with System One models, developing with JavaScript and Python SDKs, and implementing advanced AI patterns. It covers structured AI decisions, automation cookbooks, and API error handling, enabling integration, troubleshooting, and research to build reliable AI applications that make structured decisions.
- Pages
- 112
- 542
- Tokens
- ~296.5K
- Updated
- Oct 8, 2026
- Monitoring
- Active
Works with
- ChatGPT
- Claude
- Cursor
- MCP clients
Solutions
What AI Agents Can Do With This Source
Use TypeSafe AI Documentation to complete source-specific tasks, make informed decisions, and follow evidence-backed workflows.
Integrating TypeSafe AI into a JavaScript application
The JavaScript SDK documentation provides quickstart instructions and API references for installing the SDK, setting up the client, and making API calls.
Diagnosing API connection errors in the JavaScript SDK
The API reference details various error classes like APIConnectionError, providing insights into potential causes for request or response delivery failures.
Defining ChoiceQuestions for structured decisions
The API reference for ChoiceQuestion and the choice() function explains how to define questions that select between named alternatives with specified criteria.
Automating structure recovery from plain text
The "Structure recovery" cookbook details a two-pass pipeline using Noul questions to reconstruct Markdown from unformatted text without altering the original words.
Implementing composite scoring for complex judgments
The "Composite scoring" pattern guide explains how to break down complex judgments into atomic scores and combine them with weighted logic in code.
Coverage
What's Indexed from TypeSafe AI
A quick look at the main areas Bulkgrid has indexed and keeps ready for AI search, MCP, extraction, and change detection.
SDK
/sdk
65 pages
Provides comprehensive documentation for both JavaScript and Python SDKs, offering quickstart guides, API references, and details on error handling to facilitate application development.
Javascript
/sdk/javascript
52 pages
Documents the TypeSafe JavaScript and TypeScript SDK, providing detailed API references, error class definitions, and a changelog for tracking development updates.
Javascript — API
/sdk/javascript/api
50 pages
Details the JavaScript SDK's API, covering core components like error handling, client usage, and structured question types such as ChoiceQuestion and Noul.
Cookbooks
/cookbooks
18 pages
Offers practical guides and examples for applying TypeSafe AI to various automation tasks, including structure recovery, classification, and autoresearch feature discovery.
Interfaces
/sdk/javascript/api/interfaces
18 pages
Defines the interfaces for various TypeSafe AI components within the JavaScript SDK, including structured questions like ChoiceQuestion and NoulQuestion, along with their corresponding responses.
Classes
/sdk/javascript/api/classes
14 pages
Describes the classes available in the JavaScript SDK API, particularly focusing on various error types like APIConnectionError and the core TypeSafeClient for interacting with the AI.
Type Aliases
/sdk/javascript/api/type-aliases
12 pages
Lists and explains the type aliases used within the JavaScript SDK API, facilitating understanding of data structures for questions, criteria, and other core components.
Python — API
/sdk/python/api
8 pages
Details the Python SDK's API, covering client types, data structures for questions and responses, error handling, retry policies, and configuration constants for integration.
Concepts
/concepts
4 pages
Explains the foundational concepts of TypeSafe AI, including the System One model, guidance on building with it, and various example use cases across different industries.
Patterns
/patterns
4 pages
Explores advanced AI design patterns for TypeSafe AI, such as combining multiple scores, routing based on confidence, and speculative fan-out strategies for complex judgments.
Primitives
/primitives
4 pages
Introduces the fundamental building blocks of TypeSafe AI, including structured question types like Choice, Noul, and Score, and how they handle JSON structures for advanced use.
Advantages
Why Use Bulkgrid for AI-Ready Sources
Bulkgrid turns important websites into fresh, queryable, AI-ready sources without making teams maintain the infrastructure.
Spend Fewer Tokens
Retrieve the right source context instead of sending large pages, pasted docs, or noisy search results into every prompt.
Improve Answer Accuracy
Ground responses in current, source-backed evidence to reduce hallucinations and avoid relying on stale model memory.
Use Fresh Source Data
Keep important websites automatically recrawled and updated so AI is not working from old content.
Query Instantly
Search already-indexed source content immediately instead of crawling pages at question time.
No Infrastructure Maintenance
Bulkgrid handles crawling, rendering, retries, storage, indexing, recrawls, and change detection.
Change Detection
See when pages are added, removed, updated, or fail so teams can react without checking manually.
Give Your Agents Up-to-Date Knowledge from TypeSafe AI
Search and use knowledge from TypeSafe AI in ChatGPT, Claude, Cursor, and any MCP-compatible client.