Jev directory
CommunityJev Integrations & Adapters
AI SDK TypeSafe Provider
Official AI SDK provider package for TypeSafe AI. Through its decisionModel, used with the AI SDK's experimental_decide, it decides answers to Choice, Score, and Boolean questions using TypeSafe's System One models such as jev-latest.
- Reading time
- 1 min
- Published by
- Community
- Added
- Oct 2026
Highlights
- Install with npm i ai @ai-sdk/typesafe-ai
- Exposes typeSafeAi.decisionModel('jev-latest') for use with experimental_decide
- Questions are declared with choice, score, or boolean types plus instructions and criteria
- Decisions come back on result.answers
- Designed for deciding typed answers rather than generating text
Questions about this entry
Short answers drawn from the linked project or page.
Install it alongside the AI SDK itself: npm i ai @ai-sdk/typesafe-ai. You then import typeSafeAi from '@ai-sdk/typesafe-ai' and experimental_decide from 'ai'.
experimental_decide. You pass it a model, a state string, and a map of questions, and it returns typed answers for those questions.
The provider decides answers to Choice, Score, and Boolean questions. In the documented example a department question uses type 'choice' and a requestsRefund question uses type 'boolean'.
Pass a System One model to typeSafeAi.decisionModel, as in typeSafeAi.decisionModel('jev-latest') shown in the package's usage example.
The documented example logs result.answers, which holds the decided answers for each question named in the questions object.
Yes. Each question can carry instructions, and choice questions can carry a criteria map, such as billing for charges and refunds and support for other requests.
Yes. The install command brings in both ai and @ai-sdk/typesafe-ai, and the usage example imports experimental_decide from 'ai' alongside the typeSafeAi provider.
More Jev resources like this
TypeScript service and library that places Jev between a PBX such as Asterisk or FreePBX and call destinations, classifying caller intent into predefined routes. Each route maps to an explicit PBX action, with low-confidence or unhandled calls sent to a fallback route.
Jev Integrations & Adapters#typescript#routing#github
Communityrayanweragala/jev-call-router
n8n community node that sends a state and a map of Noul, Choice, and Score questions to the System One endpoint in one request, returning calibrated probabilities and confidence for workflow branching.
Jev Integrations & Adapters#github#typescript#guardrails
Communityzampierid4p/n8n-nodes-typesafe
Magento 2 module giving a store one shared Jev (TypeSafe) client: encrypted API key, pinned model version, monthly token budget with usage and cost, a response cache, and retries with backoff and a circuit breaker. Adds no storefront behaviour on its own.
Jev Integrations & Adapters#github#calibration#classification
Communitydmlab-work/module-typesafe-core
A contextual brand-suitability scorer that judges each feed item against an advertiser's own versioned policy, returning a category, an exclusion flag, and a confidence for routing borderline items to review.
Jev Integrations & Adapters#community#typescript#advertising
Community
Open-source TypeScript agentic operating system: a core runtime, app for web and mobile, CLI, cloud services and first-party plugins. The directory's curated summary reports it integrates TypeSafe System One decision services for sub-100ms intent classification, action dispatching and confidence-gated tool execution.
Jev Integrations & Adapters#community#agents#framework
Community
elizaOS/elizaTypeScript
elizaOS/elizaTypeScriptThe Vercel AI SDK provider for OpenRouter, including openrouter.decisionModel for typesafe/jev-1.13 evaluations with providerMetadata answers.
Jev Integrations & Adapters#community#typescript#openrouter
From the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.