# agentsite-static-version=092c25b7-87c9-4273-bb86-03206511edc3
# typesafe.ai

> AI-optimized mirror of typesafe.ai containing 244 pages totalling 153,696 words of clean markdown content, structured data, and semantic HTML. Original source: https://typesafe.ai. Last updated: 2026-11-11T01:03:32.097Z.

Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG).

## Homepage

- [Workflow evals](/content/evals/index.html) (1,098 words)
- [Introduction - TypeSafe AI](/content/docs/index.html): Jev is TypeSafe's flagship model and the first System One model. Send state and typed questions; get structured answers your code can use directly. (429 words)
- [Trust Center - Typesafe.ai](/content/trust/index.html): Structured Artificial Intelligence

A new class of AI models which enable a reliable way to embed composable intelligence into traditional software. (266 words)
- [Home - TypeSafe AI](/content/site-root.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access., Design production-ready web components with shaders, media, 3D, and interactive motion in a visual canvas. (430 words)

## Articles & Blog Posts

- [Don’t Call It a Decision Model - TypeSafe AI Blog](/content/blog/dont-call-it-a-decision-model/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (1,778 words)
- [Jev: For the builders AND dreamers - TypeSafe AI Blog](/content/blog/jev-1-14/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (857 words)
- [TypeSafe A raises Series AI - TypeSafe AI Blog](/content/blog/series-ai/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (217 words)
- [Accessibility - TypeSafe AI](/content/legal/accessibility/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (67 words)
- [How Jack & Jill Cut Candidate Screening Costs By 88% With Jev - TypeSafe AI Blog](/content/blog/jack-jill-jev-case-study/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (978 words)
- [TypeSafe AI](/content/enterprise/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access., Design production-ready web components with shaders, media, 3D, and interactive motion in a visual canvas., Design production-ready web components with shaders, media, 3D, and interactive motion in a visual canvas. (228 words)
- [TypeSafe AI](/content/blog/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (239 words)
- [Acceptable Use Policy - TypeSafe AI](/content/legal/acceptable-use-policy/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (804 words)
- [Function calling - TypeSafe AI](/content/docs/cookbooks/function_calling/index.html): Turns natural-language trading requests into calls to ordinary typed functions by mapping function names and closed-set arguments to confidence-aware TypeSafe questions. (1,932 words)
- [Guardrails for LLMs - TypeSafe AI](/content/docs/cookbooks/llm_guardrails/index.html): Screen every message going into and out of an LLM app with one TypeSafe request, describing possible hazards ('is this a jailbreak attempt?') and scoring severity ('how much harm would complying do?'). Threshold the probabilities it hands back and you decide whether to pass, review, block, or route a message to support. (2,244 words)
- [Date extraction - TypeSafe AI](/content/docs/cookbooks/date_extraction_cookbook/index.html): Extracts absolute and relative dates by asking TypeSafe for the parts named in a document, then resolving and validating them in code with confidence-based review. (1,712 words)
- [Re-ranking - TypeSafe AI](/content/docs/cookbooks/rerank_typesafe/index.html): Builds 30-passage BM25 shortlists for 40 CLERC legal queries, then uses one TypeSafe question per query-candidate pair to raise top-1 accuracy from 5% to 18% and top-10 accuracy from 38% to 62%. (2,265 words)
- [Self-consistency: choices - TypeSafe AI](/content/docs/cookbooks/consistency_choice_cookbook/index.html): Add an uncertain outcome to moderation decisions and compare label agreement with the share of automatic actions. (2,098 words)
- [Pre-parsed value extraction - TypeSafe AI](/content/docs/cookbooks/pre_parsed_value_extraction_cookbook/index.html): Uses regexes to find candidate emails, phone numbers, and amounts, then has TypeSafe select the requested span so code can normalize a verbatim value. (1,360 words)
- [Hierarchical classification - TypeSafe AI](/content/docs/cookbooks/hierarchical_classification/index.html): Classifies documents through deep patent, retail product, biomedical, and source-code hierarchies using parallel beam search over TypeSafe Choice probabilities. (802 words)
- [docs/cookbooks/hierarchical_classification-md.html](/content/docs/cookbooks/hierarchical_classification-md.html) (2,468 words)
- [How to build with TypeSafe - TypeSafe AI](/content/docs/concepts/how-to-build-with-system-one/index.html): Design AI-powered software by keeping code in control and giving System One narrow, structured decisions. (2,032 words)
- [Skill suggestion - TypeSafe AI](/content/docs/cookbooks/skill_suggestion/index.html): Picks at most one skill for an agent turn out of the 182 in Nous Research's Hermes catalog: one TypeSafe request ranks every skill and asks whether the turn needs one at all, a second reads the top three properly and can reject all of them. The winner's name goes into a single line of the agent's system prompt, and both the wrong skills it loads and the ones it loads when nothing fits drop by more than half. (3,498 words)
- [docs/cookbooks/rerank_typesafe-md.html](/content/docs/cookbooks/rerank_typesafe-md.html) (2,748 words)
- [Knowledge graph entity alignment - TypeSafe AI](/content/docs/cookbooks/entity_alignment/index.html): Decides which of 450 candidate pairs from two beer catalogues describe the same product. One TypeSafe Score question carries the whole decision, because its three levels are the three things you can do with a pair: merge it, leave it unlinked, or hand it to a curator. There is no threshold to fit, and three Noul questions ride along in the same request to tell the curator which field the two sources disagree on. (704 words)
- [docs/cookbooks/function_calling-md.html](/content/docs/cookbooks/function_calling-md.html) (1,924 words)
- [Double-checking citations - TypeSafe AI](/content/docs/cookbooks/citation_check/index.html): Catch wrong or hallucinated citations by checking against the source document. One TypeSafe Choice question decides whether the quote's context supports the claim, and its confidence can flag the citation for human review. (1,148 words)
- [Classification using confidence - TypeSafe AI](/content/docs/cookbooks/classification_using_confidence/index.html): Classify SEC annual reports into 75 industry groups with one Choice each, then read the answer's own confidence to decide whether to report that group or the broader division above it. (1,875 words)
- [docs/cookbooks/date_extraction_cookbook-md.html](/content/docs/cookbooks/date_extraction_cookbook-md.html) (2,061 words)
- [Classifying RAG passages - TypeSafe AI](/content/docs/cookbooks/classifying_rag_passages/index.html): Score each retrieved passage with one TypeSafe request, then decide in code which ones reach the answering model. For example, keep and flag ones that contradict the question, and drop ones carrying a hidden instruction or prompt injection. (2,982 words)
- [docs/cookbooks/classification_using_confidence-md.html](/content/docs/cookbooks/classification_using_confidence-md.html) (2,154 words)
- [docs/cookbooks/llm_guardrails-md.html](/content/docs/cookbooks/llm_guardrails-md.html) (1,563 words)
- [docs/cookbooks/citation_check-md.html](/content/docs/cookbooks/citation_check-md.html) (1,787 words)
- [Security Incidents](/content/evals/security_incidents-html.html) (437 words)
- [SDE cascade - TypeSafe AI](/content/docs/cookbooks/sde_cascade/index.html): Uses a 2-stage structured-data-extraction cascade (mini → verify → reasoning) to get most of the quality of a big reasoning model at a fraction of the cost. (1,191 words)
- [docs/primitives/choice-md.html](/content/docs/primitives/choice-md.html) (1,339 words)
- [docs/cookbooks/pre_parsed_value_extraction_cookbook-md.html](/content/docs/cookbooks/pre_parsed_value_extraction_cookbook-md.html) (1,213 words)
- [docs/patterns/fan-out-md.html](/content/docs/patterns/fan-out-md.html) (1,301 words)
- [docs/cookbooks/entity_alignment-md.html](/content/docs/cookbooks/entity_alignment-md.html) (2,069 words)
- [docs/cookbooks/sde_cascade-md.html](/content/docs/cookbooks/sde_cascade-md.html) (2,553 words)
- [docs/sdk/python/api/types/responses-md.html](/content/docs/sdk/python/api/types/responses-md.html) (759 words)
- [Self-consistency: nouls - TypeSafe AI](/content/docs/cookbooks/consistency_noul_cookbook/index.html): Route uncertain probabilities to human review while keeping the underlying noul values visible. (2,015 words)
- [docs/patterns/composite-scoring-md.html](/content/docs/patterns/composite-scoring-md.html) (652 words)
- [Async client - TypeSafe AI](/content/docs/sdk/python/api/clients/async/index.html): Use AsyncTypeSafeClient to ask questions, list models, and configure asynchronous TypeSafe API requests. (956 words)
- [Agent Trace Observability](/content/evals/agent_trace_observability-html.html) (550 words)
- [Invoice Processing](/content/evals/invoice_processing-html.html) (377 words)
- [Customer Service](/content/evals/customer_service-html.html) (519 words)
- [Sync client - TypeSafe AI](/content/docs/sdk/python/api/clients/sync/index.html): Use TypeSafeClient to ask questions, list models, and configure synchronous TypeSafe API requests. (1,036 words)
- [docs/patterns/intent-routing-md.html](/content/docs/patterns/intent-routing-md.html) (1,213 words)
- [Answers and responses - TypeSafe AI](/content/docs/sdk/python/api/types/responses/index.html): Read answers, confidence scores, token usage, and available models returned by the TypeSafe API. (1,210 words)
- [API reference - TypeSafe AI](/content/docs/sdk/javascript/api/index.html) (56 words)
- [docs/cookbooks/consistency_choice_cookbook-md.html](/content/docs/cookbooks/consistency_choice_cookbook-md.html) (969 words)
- [docs/api-md.html](/content/docs/api-md.html) (876 words)
- [docs/patterns/confidence-routing-md.html](/content/docs/patterns/confidence-routing-md.html) (1,122 words)
- [docs/primitives/advanced-md.html](/content/docs/primitives/advanced-md.html) (1,649 words)
- [Usage - TypeSafe AI](/content/docs/sdk/python/usage/index.html): Guides and patterns for working with the TypeSafe Python SDK. (630 words)
- [Retries - TypeSafe AI](/content/docs/sdk/python/api/retries/index.html): Configure retries with RetryPolicy — attempt count, retryable statuses, backoff, and retry headers handling. (316 words)
- [docs/model-jaggedness/jev-1-13-md.html](/content/docs/model-jaggedness/jev-1-13-md.html) (1,516 words)
- [Jev 1.13 jaggedness - TypeSafe AI](/content/docs/model-jaggedness/jev-1-13.html): Jev isn't perfect. Here are some jagged edges we are aware of with jev-1.13. Many of these will be fixed in later versions. (1,447 words)
- [Class: TypeSafeClient - TypeSafe AI](/content/docs/sdk/javascript/api/classes/typesafeclient/index.html) (264 words)
- [Structure recovery - TypeSafe AI](/content/docs/cookbooks/autoformat/index.html): Reconstructs Markdown from plain text that lost its formatting in two requests: one stitches hard-wrapped lines back together, one classifies every block (heading, list, code, callout) with companion questions read only when relevant. (3,605 words)
- [Parallel questions - TypeSafe AI](/content/docs/cookbooks/parallel_questions/index.html): Runs a 13-question regulatory briefing over the GDPR Wikipedia article, showing that batching every question into one TypeSafe call is 12.2x cheaper and 10.0x faster with no change in answers. (1,744 words)
- [docs/cookbooks/consistency_noul_cookbook-md.html](/content/docs/cookbooks/consistency_noul_cookbook-md.html) (1,221 words)
- [Class: UnprocessableEntityError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/unprocessableentityerror/index.html) (119 words)
- [Interface: TypeSafeClientConfig - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/typesafeclientconfig/index.html) (163 words)
- [Class: InternalServerError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/internalservererror/index.html) (143 words)
- [Function: noul() - TypeSafe AI](/content/docs/sdk/javascript/api/functions/noul/index.html) (102 words)
- [Function: choice() - TypeSafe AI](/content/docs/sdk/javascript/api/functions/choice/index.html) (66 words)
- [Class: RateLimitError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/ratelimiterror/index.html) (138 words)
- [Class: PermissionDeniedError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/permissiondeniederror/index.html) (118 words)
- [Interface: RetryPolicy - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/retrypolicy/index.html) (139 words)
- [docs/agent-skill-md.html](/content/docs/agent-skill-md.html) (723 words)
- [Type Alias: ResultFor<T> - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/resultfor/index.html) (60 words)
- [Interface: ChoiceResponse<T> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/choiceresponse/index.html) (51 words)
- [Variable: ENV - TypeSafe AI](/content/docs/sdk/javascript/api/variables/env/index.html) (83 words)
- [Class: NotFoundError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/notfounderror/index.html) (114 words)
- [docs/concepts/how-to-build-with-system-one-md.html](/content/docs/concepts/how-to-build-with-system-one-md.html) (2,328 words)
- [Type Alias: ScoreLegend<T> - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/scorelegend/index.html) (41 words)
- [Class: APIPromise<T> - TypeSafe AI](/content/docs/sdk/javascript/api/classes/apipromise/index.html) (337 words)
- [Type Alias: ScoreOf<T> - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/scoreof/index.html) (52 words)
- [Variable: VERSION - TypeSafe AI](/content/docs/sdk/javascript/api/variables/version/index.html) (25 words)
- [Line-by-line search - TypeSafe AI](/content/docs/cookbooks/semantic_find/index.html): Build semantic search for GitHub's Terms of Service. In one request, score 218 line ids against a plain-language query with a Choice question, and use a Noul question to check whether the document contains an answer. (1,327 words)
- [Class: AuthenticationError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/authenticationerror/index.html) (122 words)
- [API reference - TypeSafe AI](/content/docs/sdk/python/api/index.html): Python clients for the TypeSafe AI API (20 words)
- [docs/models-md.html](/content/docs/models-md.html) (701 words)
- [docs/primitives-md.html](/content/docs/primitives-md.html) (401 words)
- [Type Alias: EnvVar - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/envvar/index.html) (27 words)
- [Type Alias: LogLevel - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/loglevel/index.html) (37 words)
- [Interface: ScoreResponse<T> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/scoreresponse/index.html) (60 words)
- [Interface: Logger - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/logger/index.html) (56 words)
- [Primitives (Questions) - TypeSafe AI](/content/docs/primitives/index.html): The three TypeSafe question types (Choice, Score, Noul), the typed answers they return, how to choose between them, and how to ask several at once. (2,054 words)
- [Type Alias: Fetch - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/fetch/index.html) (42 words)
- [Interface: SystemOneResult<Q> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/systemoneresult/index.html) (49 words)
- [Type Alias: Description - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/description/index.html) (32 words)
- [Type Alias: JsonValue - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/jsonvalue/index.html) (42 words)
- [Interface: WithResponse<T> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/withresponse/index.html) (37 words)
- [docs/sdk/python/usage-md.html](/content/docs/sdk/python/usage-md.html) (560 words)
- [Interface: Usage - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/usage/index.html) (44 words)
- [Interface: NoulQuestion - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/noulquestion/index.html) (104 words)
- [docs/sdk/python/api/types/questions-md.html](/content/docs/sdk/python/api/types/questions-md.html) (240 words)
- [Exceptions - TypeSafe AI](/content/docs/sdk/python/api/exceptions/index.html): Handle TypeSafe API errors, rate limits, connection failures, and timeouts. (251 words)
- [docs/introduction/machine-learning-primer-md.html](/content/docs/introduction/machine-learning-primer-md.html) (503 words)
- [Class: APIError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/apierror/index.html) (114 words)
- [Questions - TypeSafe AI](/content/docs/sdk/python/api/types/questions/index.html): Provide state and ask yes/no, choice, and score questions using objects or dictionaries. (282 words)
- [Interface: ChoiceQuestion<T> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/choicequestion/index.html) (37 words)
- [docs/cookbooks/classifying_rag_passages-md.html](/content/docs/cookbooks/classifying_rag_passages-md.html) (1,426 words)
- [Interface: NoulResponse - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/noulresponse/index.html) (18 words)
- [Interface: ScoreQuestion<T> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/scorequestion/index.html) (47 words)
- [docs/introduction/quickstart-md.html](/content/docs/introduction/quickstart-md.html) (588 words)
- [Class: APIUserAbortError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/apiuseraborterror/index.html) (41 words)
- [Interface: SystemOneRequestPayload - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/systemonerequestpayload/index.html) (59 words)
- [Changelog - TypeSafe AI](/content/docs/sdk/python/changelog/index.html): Python clients for the TypeSafe AI API (127 words)
- [docs/sdk/python/api/exceptions-md.html](/content/docs/sdk/python/api/exceptions-md.html) (271 words)
- [Function: score() - TypeSafe AI](/content/docs/sdk/javascript/api/functions/score/index.html) (70 words)
- [docs/primitives/noul-md.html](/content/docs/primitives/noul-md.html) (322 words)
- [Interface: SystemOneRequest<Q> - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/systemonerequest/index.html) (86 words)
- [Type Alias: EntryType - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/entrytype/index.html) (48 words)
- [docs/sdk/python/api/clients/sync-md.html](/content/docs/sdk/python/api/clients/sync-md.html) (122 words)
- [Interface: Models - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/models/index.html) (26 words)
- [Class: APITimeoutError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/apitimeouterror/index.html) (60 words)
- [docs/sdk/python/api/clients/async-md.html](/content/docs/sdk/python/api/clients/async-md.html) (581 words)
- [Constants - TypeSafe AI](/content/docs/sdk/python/api/constants/index.html): Default settings and environment variable names for the TypeSafe Python SDK. (82 words)
- [Common types - TypeSafe AI](/content/docs/sdk/python/api/types/common/index.html): Common types for TypeSafe API SDK. (59 words)
- [docs/primitives/score-md.html](/content/docs/primitives/score-md.html) (1,491 words)
- [Interface: ModelCard - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/modelcard/index.html) (18 words)
- [Type Alias: ScoreCriteria - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/scorecriteria/index.html) (41 words)
- [docs/cookbooks/autoformat-md.html](/content/docs/cookbooks/autoformat-md.html) (228 words)
- [docs/demos/smart-home-md.html](/content/docs/demos/smart-home-md.html) (564 words)
- [Interface: RequestOptions - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/requestoptions/index.html) (52 words)
- [Class: TypeSafeError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/typesafeerror/index.html) (47 words)
- [Class: APIConnectionError - TypeSafe AI](/content/docs/sdk/javascript/api/classes/apiconnectionerror/index.html) (24 words)
- [Type Alias: ChoiceCriteria - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/choicecriteria/index.html) (38 words)
- [Changelog - TypeSafe AI](/content/docs/sdk/javascript/changelog/index.html) (36 words)
- [AI primer - TypeSafe AI](/content/docs/introduction/machine-learning-primer/index.html): Why TypeSafe trains decision models with calibrated probabilities instead of optimizing for generated text. (514 words)
- [docs/sdk/javascript/api/classes/typesafeclient-md.html](/content/docs/sdk/javascript/api/classes/typesafeclient-md.html) (265 words)
- [API reference - TypeSafe AI](/content/docs/api/index.html): Full HTTP API reference for the TypeSafe evaluation endpoint. (779 words)
- [Type Alias: Question - TypeSafe AI](/content/docs/sdk/javascript/api/type-aliases/question/index.html) (36 words)
- [Example use cases - TypeSafe AI](/content/docs/concepts/use-case-map/index.html): Explore TypeSafe use cases by industry and turn promising ideas into software workflows. (911 words)
- [docs/sdk/javascript/api/functions/noul-md.html](/content/docs/sdk/javascript/api/functions/noul-md.html) (83 words)
- [Autoresearch feature discovery - TypeSafe AI](/content/docs/cookbooks/autoresearch_feature_discovery/index.html): Runs an autoresearch loop that proposes TypeSafe questions, converts free text into numeric features, and uses model errors to improve a supervised CatBoost regressor. (7,496 words)
- [Interface: Questions - TypeSafe AI](/content/docs/sdk/javascript/api/interfaces/questions/index.html) (34 words)
- [docs/sdk/javascript/api/classes/internalservererror-md.html](/content/docs/sdk/javascript/api/classes/internalservererror-md.html) (123 words)
- [docs/sdk/javascript/api/classes/apipromise-md.html](/content/docs/sdk/javascript/api/classes/apipromise-md.html) (322 words)
- [docs/sdk/javascript/api/classes/unprocessableentityerror-md.html](/content/docs/sdk/javascript/api/classes/unprocessableentityerror-md.html) (119 words)
- [docs/sdk/javascript/api-md.html](/content/docs/sdk/javascript/api-md.html) (58 words)
- [docs/concepts/state-md.html](/content/docs/concepts/state-md.html) (376 words)
- [Choice - TypeSafe AI](/content/docs/primitives/choice/index.html): A Choice is a System One question type for selecting one option from a defined set. The answer includes the selected option, a probability for each option, and confidence. (731 words)
- [docs/confidence-md.html](/content/docs/confidence-md.html) (565 words)
- [docs/concepts/system-one-md.html](/content/docs/concepts/system-one-md.html) (511 words)
- [docs/sdk/javascript/api/interfaces/retrypolicy-md.html](/content/docs/sdk/javascript/api/interfaces/retrypolicy-md.html) (141 words)
- [docs/sdk/javascript/api/classes/permissiondeniederror-md.html](/content/docs/sdk/javascript/api/classes/permissiondeniederror-md.html) (119 words)
- [docs/sdk/javascript/api/classes/ratelimiterror-md.html](/content/docs/sdk/javascript/api/classes/ratelimiterror-md.html) (138 words)
- [docs/introduction-md.html](/content/docs/introduction-md.html) (424 words)
- [docs/sdk/python/api/retries-md.html](/content/docs/sdk/python/api/retries-md.html) (280 words)
- [docs/sdk/python/api/types/common-md.html](/content/docs/sdk/python/api/types/common-md.html) (59 words)
- [docs/sdk/javascript/api/interfaces/typesafeclientconfig-md.html](/content/docs/sdk/javascript/api/interfaces/typesafeclientconfig-md.html) (165 words)
- [docs/sdk/javascript/api/classes/authenticationerror-md.html](/content/docs/sdk/javascript/api/classes/authenticationerror-md.html) (118 words)
- [docs/sdk/javascript/api/classes/notfounderror-md.html](/content/docs/sdk/javascript/api/classes/notfounderror-md.html) (121 words)
- [docs/legal-md.html](/content/docs/legal-md.html) (112 words)
- [docs/sdk/python-md.html](/content/docs/sdk/python-md.html) (204 words)
- [docs/sdk/javascript/api/classes/badrequesterror-md.html](/content/docs/sdk/javascript/api/classes/badrequesterror-md.html) (120 words)
- [docs/sdk/javascript/api/classes/apierror-md.html](/content/docs/sdk/javascript/api/classes/apierror-md.html) (114 words)
- [docs/sdk/javascript/api/functions/choice-md.html](/content/docs/sdk/javascript/api/functions/choice-md.html) (47 words)
- [Agent skill - TypeSafe AI](/content/docs/agent-skill/index.html): Drop-in skill for Claude Code, Codex, and other agent environments. (731 words)
- [docs/sdk/javascript/api/functions/score-md.html](/content/docs/sdk/javascript/api/functions/score-md.html) (51 words)
- [Advanced: structure - TypeSafe AI](/content/docs/primitives/advanced/index.html): Instructions, Choice options, Score levels, and Noul criteria all accept JSON structure. (334 words)
- [docs/sdk/javascript/api/variables/env-md.html](/content/docs/sdk/javascript/api/variables/env-md.html) (65 words)
- [docs/sdk/javascript/api/type-aliases/scorelegend-md.html](/content/docs/sdk/javascript/api/type-aliases/scorelegend-md.html) (24 words)
- [docs/cookbooks/autoresearch_feature_discovery-md.html](/content/docs/cookbooks/autoresearch_feature_discovery-md.html) (7,508 words)
- [docs/sdk/javascript/api/type-aliases/scoreof-md.html](/content/docs/sdk/javascript/api/type-aliases/scoreof-md.html) (35 words)
- [TypeSafe Python SDK - TypeSafe AI](/content/docs/sdk/python/index.html): Install the TypeSafe Python SDK and get started with asynchronous or synchronous API calls. (169 words)
- [docs/sdk/javascript/api/classes/apitimeouterror-md.html](/content/docs/sdk/javascript/api/classes/apitimeouterror-md.html) (41 words)
- [docs/sdk/python/api/constants-md.html](/content/docs/sdk/python/api/constants-md.html) (114 words)
- [docs/cookbooks/semantic_find-md.html](/content/docs/cookbooks/semantic_find-md.html) (1,349 words)
- [docs/sdk/python/changelog-md.html](/content/docs/sdk/python/changelog-md.html) (128 words)
- [docs/sdk/javascript/api/type-aliases/resultfor-md.html](/content/docs/sdk/javascript/api/type-aliases/resultfor-md.html) (43 words)
- [docs/sdk/javascript/api/type-aliases/fetch-md.html](/content/docs/sdk/javascript/api/type-aliases/fetch-md.html) (25 words)
- [docs/sdk/javascript-md.html](/content/docs/sdk/javascript-md.html) (115 words)
- [docs/sdk/javascript/api/variables/version-md.html](/content/docs/sdk/javascript/api/variables/version-md.html) (27 words)
- [docs/sdk/javascript/api/variables/log_levels-md.html](/content/docs/sdk/javascript/api/variables/log_levels-md.html) (14 words)
- [docs/sdk/javascript/api/type-aliases/question-md.html](/content/docs/sdk/javascript/api/type-aliases/question-md.html) (19 words)
- [docs/sdk/javascript/api/type-aliases/jsonvalue-md.html](/content/docs/sdk/javascript/api/type-aliases/jsonvalue-md.html) (25 words)
- [docs/cookbooks/parallel_questions-md.html](/content/docs/cookbooks/parallel_questions-md.html) (1,756 words)
- [Score - TypeSafe AI](/content/docs/primitives/score/index.html): A Score is a System One question type for rating content against ordered, descriptive levels. The answer includes a score, a probability for each level, and confidence. (665 words)
- [docs/sdk/javascript/api/type-aliases/scorecriteria-md.html](/content/docs/sdk/javascript/api/type-aliases/scorecriteria-md.html) (24 words)
- [docs/sdk/javascript/api/interfaces/systemonerequestpayload-md.html](/content/docs/sdk/javascript/api/interfaces/systemonerequestpayload-md.html) (58 words)
- [docs/sdk/javascript/api/interfaces/noulquestion-md.html](/content/docs/sdk/javascript/api/interfaces/noulquestion-md.html) (81 words)
- [Models - TypeSafe AI](/content/docs/models/index.html) (771 words)
- [docs/sdk/javascript/api/interfaces/scoreresponse-md.html](/content/docs/sdk/javascript/api/interfaces/scoreresponse-md.html) (70 words)
- [docs/sdk/javascript/api/type-aliases/loglevel-md.html](/content/docs/sdk/javascript/api/type-aliases/loglevel-md.html) (20 words)
- [docs/patterns-md.html](/content/docs/patterns-md.html) (129 words)
- [docs/sdk/javascript/api/interfaces/logger-md.html](/content/docs/sdk/javascript/api/interfaces/logger-md.html) (58 words)
- [docs/sdk/javascript/api/type-aliases/entrytype-md.html](/content/docs/sdk/javascript/api/type-aliases/entrytype-md.html) (18 words)
- [docs/sdk/javascript/api/interfaces/withresponse-md.html](/content/docs/sdk/javascript/api/interfaces/withresponse-md.html) (48 words)
- [docs/sdk/javascript/changelog-md.html](/content/docs/sdk/javascript/changelog-md.html) (37 words)
- [docs/sdk/javascript/api/interfaces/systemoneresult-md.html](/content/docs/sdk/javascript/api/interfaces/systemoneresult-md.html) (61 words)
- [docs/sdk/javascript/api/interfaces/scorequestion-md.html](/content/docs/sdk/javascript/api/interfaces/scorequestion-md.html) (49 words)
- [docs/sdk/javascript/api/type-aliases/envvar-md.html](/content/docs/sdk/javascript/api/type-aliases/envvar-md.html) (30 words)
- [Confidence - TypeSafe AI](/content/docs/confidence/index.html): How TypeSafe reports certainty, how it differs from probability, and how to use it to control system behavior. (674 words)
- [docs/sdk/javascript/api/interfaces/models-md.html](/content/docs/sdk/javascript/api/interfaces/models-md.html) (26 words)
- [docs/sdk/javascript/api/interfaces/systemonerequest-md.html](/content/docs/sdk/javascript/api/interfaces/systemonerequest-md.html) (68 words)
- [docs/sdk/javascript/api/classes/apiuseraborterror-md.html](/content/docs/sdk/javascript/api/classes/apiuseraborterror-md.html) (31 words)
- [docs/sdk/javascript/api/interfaces/requestoptions-md.html](/content/docs/sdk/javascript/api/interfaces/requestoptions-md.html) (60 words)
- [docs/sdk/javascript/api/interfaces/choiceresponse-md.html](/content/docs/sdk/javascript/api/interfaces/choiceresponse-md.html) (59 words)
- [docs/sdk/javascript/api/classes/apiconnectionerror-md.html](/content/docs/sdk/javascript/api/classes/apiconnectionerror-md.html) (36 words)
- [docs/sdk/javascript/api/classes/typesafeerror-md.html](/content/docs/sdk/javascript/api/classes/typesafeerror-md.html) (29 words)
- [docs/sdk/javascript/api/interfaces/choicequestion-md.html](/content/docs/sdk/javascript/api/interfaces/choicequestion-md.html) (46 words)
- [docs/sdk/javascript/api/type-aliases/description-md.html](/content/docs/sdk/javascript/api/type-aliases/description-md.html) (15 words)
- [Trust Center - Typesafe.ai](/content/trust/controls/index.html): Structured Artificial Intelligence

A new class of AI models which enable a reliable way to embed composable intelligence into traditional software. (902 words)
- [docs/sdk-md.html](/content/docs/sdk-md.html) (84 words)
- [docs/sdk/javascript/api/interfaces/modelcard-md.html](/content/docs/sdk/javascript/api/interfaces/modelcard-md.html) (20 words)
- [docs/sdk/javascript/api/type-aliases/choicecriteria-md.html](/content/docs/sdk/javascript/api/type-aliases/choicecriteria-md.html) (21 words)
- [JavaScript SDK - TypeSafe AI](/content/docs/sdk/javascript/index.html) (106 words)
- [docs/sdk/python/api-md.html](/content/docs/sdk/python/api-md.html) (22 words)
- [docs/sdk/javascript/api/interfaces/usage-md.html](/content/docs/sdk/javascript/api/interfaces/usage-md.html) (26 words)
- [docs/sdk/javascript/api/interfaces/noulresponse-md.html](/content/docs/sdk/javascript/api/interfaces/noulresponse-md.html) (23 words)
- [docs/sdk/javascript/api/interfaces/questions-md.html](/content/docs/sdk/javascript/api/interfaces/questions-md.html) (16 words)
- [State - TypeSafe AI](/content/docs/concepts/state/index.html): What state is, how to structure it, and how to give a System One model the context it needs. (461 words)
- [System One - TypeSafe AI](/content/docs/concepts/system-one/index.html): System One models make fast, structured decisions for software. Jev is TypeSafe's flagship model and the first System One model. (493 words)
- [Noul - TypeSafe AI](/content/docs/primitives/noul/index.html): A Noul question asks the model to evaluate a yes/no question and return the probability that the answer is yes. (453 words)
- [Composite scoring - TypeSafe AI](/content/docs/patterns/composite-scoring/index.html): Break a complex judgment into atomic scores, combine with weights you control in code. (465 words)
- [Smart home assistant demo - TypeSafe AI](/content/docs/demos/smart-home/index.html): Demo code: a smart home assistant that uses TypeSafe to evaluate user requests., Use Loom to record quick videos of your screen and cam. Explain anything clearly and easily – and skip the meeting. An essential tool for hybrid workplaces. (575 words)
- [Speculative fan-out - TypeSafe AI](/content/docs/patterns/fan-out/index.html): Send many questions in a single call, including speculative ones, and let your code decide what's relevant. (459 words)
- [docs/llms-txt.html](/content/docs/llms-txt.html) (759 words)
- [Intent routing - TypeSafe AI](/content/docs/patterns/intent-routing/index.html): Classify incoming requests and route each to the optimal handler: deterministic logic, a specialist LLM, or a human. (332 words)
- [Trust Center - Typesafe.ai](/content/trust/subprocessors/index.html): Structured Artificial Intelligence

A new class of AI models which enable a reliable way to embed composable intelligence into traditional software. (130 words)
- [Confidence-gated routing - TypeSafe AI](/content/docs/patterns/confidence-routing/index.html): Use confidence as a second axis. The answer tells you what; confidence tells you whether to act. (251 words)
- [Client SDKs - TypeSafe AI](/content/docs/sdk/index.html): Install a TypeSafe client SDK and use typed questions and answers in your application. (86 words)
- [Legal - TypeSafe AI](/content/docs/legal/index.html): Legal documents and policies for TypeSafe. (67 words)
- [Patterns - TypeSafe AI](/content/docs/patterns/index.html): Architectural patterns for building systems with TypeSafe. (104 words)
- [Trust Center - Typesafe.ai](/content/trust/resources/index.html): Structured Artificial Intelligence

A new class of AI models which enable a reliable way to embed composable intelligence into traditional software. (62 words)
- [Demos - TypeSafe AI](/content/docs/demos/index.html): Interactive examples showing what's possible with TypeSafe. (51 words)
- [legal/mca/index.html](/content/legal/mca/index.html) (5,248 words)
- [Terms of use - TypeSafe AI](/content/legal/terms/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (3,442 words)
- [Data processing addendum - TypeSafe AI](/content/legal/data-processing/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (1,965 words)
- [Introducing System One Models & Jev - TypeSafe AI Blog](/content/blog/introducing-system-one-models-and-jev/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (2,119 words)
- [Privacy policy - TypeSafe AI](/content/legal/privacy-policy/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (1,579 words)
- [Manifesto - TypeSafe AI](/content/manifesto/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (735 words)
- [Lies, Damned Lies, and Benchmarks - TypeSafe AI Blog](/content/blog/antibenchmaxxing/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (797 words)
- [The Bitterest Lesson - TypeSafe AI Blog](/content/blog/bitterest-lesson/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (458 words)
- [Diogo Almeida - Founders You Should Know - TypeSafe AI Blog](/content/blog/diogo-almeida-founders-you-should-know/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (10 words)
- [AI: too good to be true, too bad to be useful | TypeSafe AI - TypeSafe AI Blog](/content/blog/ai-too-good-to-be-true-too-bad-to-be-useful-typesafe-ai.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access. (17 words)
- [robots-txt.html](/content/robots-txt.html) (6 words)

## About Pages

- [TypeSafe AI](/content/contact/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access., Design production-ready web components with shaders, media, 3D, and interactive motion in a visual canvas. (10 words)
- [Team - TypeSafe AI](/content/team/index.html): TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access., Design production-ready web components with shaders, media, 3D, and interactive motion in a visual canvas. (452 words)

## Resources

- [Full Page Index](/index.html): Browse all cached pages with rich metadata
- [About This Cache](/about.html): Methodology, technical details, and usage guidelines
- [XML Sitemap](/sitemap.xml): Machine-readable sitemap for crawler discovery
- [Robots.txt](/robots.txt): Crawler directives
- [Agentsite Network](https://agentsite.network/network.html): Public index of Agentsites and their machine-readable resources