Jev directory
CommunityJev Integrations & Adapters
Standard Agents TypeSafe Provider
First-party TypeSafe AI decision provider for Standard Agents. It implements DecisionProviderInterface with kind 'decision' and decide(), sending explicit context and multiple independent questions to Jev in one request.
- Reading time
- 2 min
- Published by
- Community
- Added
- Nov 2026
Highlights
- Implements DecisionProviderInterface with kind 'decision' and a decide() method.
- Evaluates multiple independent questions against one shared context in a single POST /v1/systemone request.
- Supports choice questions with up to 255 options, score questions with 2-10 ordered levels, and boolean questions.
- Answers return 0-1 probabilities and confidence; boolean results are never coerced to JavaScript booleans.
- Install with pnpm add @standardagents/typesafe and supply a key via TYPESAFE_AI_API_KEY.
Questions about this entry
Short answers drawn from the linked project or page.
Run pnpm add @standardagents/typesafe, then set TYPESAFE_AI_API_KEY in local .dev.vars or production secrets for BYOK. Import typesafe from the package and pass it as the provider to state.decide().
A result containing the resolved model, named answers, standard usage, and provider metadata. Choice answers carry choice, probabilities, and confidence; score answers carry score, legend, probabilities, and confidence; boolean answers carry a probability.
Choice questions accept up to 255 options, and score questions accept 2 to 10 ordered levels. Use null for a choice option that needs no extra description.
No. Boolean answers come back as a probability between 0 and 1, converted from the native noul type, and are never coerced to JavaScript booleans. Your application chooses and validates a threshold.
No. Context is explicit, as a string, JSON object, or array, and you select relevant conversation content yourself. If an answer belongs in history, call state.injectMessage() explicitly.
The context you pass plus each question's instructions, options, levels, or criteria, all evaluated in one POST /v1/systemone request. Images, audio, video, and binary data must be pre-processed into text or structured fields.
No. Provider and platform destinations come only from trusted instance-admin or deployment configuration; user or thread URL overrides cannot redirect inherited credentials. Hosted requests require a platform-owned TypeSafe credential.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
Facebook: Intake Router keeps diagnosis out of the answer space at the front desk.
FacebookClinicnotes
X: OpenRouter Decision Model Rankings
X: herdr + jev, an agent router
X: A group chat bot with decision room routing
More Jev resources like this
LangChain.js integration for TypeSafe System One models. It wraps the jev classifier as a Runnable that returns typed choice, score, and noul answers with calibrated probabilities, plus agent middleware for model routing and tool gating.
Jev Integrations & Adapters#typescript#sdk#routing
Community
Optional TypeSafe AI Jev decision provider for the Zhin chat framework. Evaluates named Choice, Score and Noul questions through a plugin-managed client, returning labels, distributions and confidence without generating chat replies or granting tool permissions.
Jev Integrations & Adapters#typescript#sdk#routing
Community
npm adapter that adds optional TypeSafe Jev system one decisions to AgentRun DSL workflows. createJevRunner turns judge, pick, sift and route steps into typed Choice, Noul and Score answers, with retries, timeouts and OpenRouter support.
Jev Integrations & Adapters#typescript#sdk#agents
Community
TypeSafe Jev adapter for TanStack AI. It exposes typed evaluate decisions over HTTP, letting TanStack AI applications route requests through Jev's System One model from JavaScript or TypeScript.
Jev Integrations & Adapters#typescript#javascript#sdk
Community
n8n community node that sends a state and typed questions to TypeSafe AI System One models, including Jev. Evaluate adds answers to each item; Route sends each item to an output based on one answer.
Jev Integrations & Adapters#typescript#routing#classification
Community
npm package that plugs TypeSafe Jev into Scruple as a decision provider. Scruple calls it for final rule decisions and for bounded collection classifications where a rule needs semantic help choosing candidates.
Jev Integrations & Adapters#typescript#routing#guardrails
Community
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.