questions: { action: choice, done: boolean, risk: boolean, }
Jev-cu
把 Computer Use 的「下一步点哪里」交给 Jev(TypeSafe System One):Jev 从界面文字候选中选元素、动作、完成度与风险,Codex Computer Use 负责读取界面与执行,本地策略门槛拦截敏感操作。只传文字,不传截图。

Every project here really calls Jev in its source · refreshed daily
questions: { action: choice, done: boolean, risk: boolean, }
把 Computer Use 的「下一步点哪里」交给 Jev(TypeSafe System One):Jev 从界面文字候选中选元素、动作、完成度与风险,Codex Computer Use 负责读取界面与执行,本地策略门槛拦截敏感操作。只传文字,不传截图。
questions: { urgency: score, }
A small, type-safe client for asking AI questions about your data, powered by TypeSafe Jev.
questions: { urgent: boolean, }
Ruby client for decision models such as Typesafe Jev
questions: { injection: boolean, hidden: boolean, exfiltration: boolean, // +3 }
Agent-ergonomic CLI for TypeSafe's Jev: fast calibrated judgments (pick, rate, check, rank, triage, guard) from the shell
questions: { movement: choice, start_jump: boolean, ceiling_hop: boolean, // +1 }
A uv-managed Python prototype that plays NES Super Mario Bros. (level 1-1 by default).
questions: { action: choice, in_danger: boolean, threat: score, }
Experiments in driving real-time games with TypeSafe Jev: a fast
questions: { billing: boolean, urgency: score, }
A small async LangGraph workflow that sends a mocked email to TypeSafe's Jev model,
questions: { is_spam: boolean, }
Run a JSON Schema through TypeSafe's Jev API, and get JSON back.
questions: { odd: boolean, }
Check whether a number is odd, using a System One model, with a calibrated probability.
questions: { department: boolean, frustration: boolean, }
Ruby client for typesafe.ai
questions: { delivery: boolean, department: choice, urgency: score, }
Adapt local language models into Jev-compatible structured decision engines with Choice, Score, and Noul outputs powered by prefill-only binary inference.
questions: { department: choice, frustration: score, is_urgent: boolean, }
DSH bundle that registers jev_ask for TypeSafe Jev noul, choice, and score answers.
questions: { urgent: boolean, }
The decision layer for your Rails app. A Rails-native wrapper around TypeSafe's Jev System One API: typed, calibrated decisions in your control flow.
questions: { requires_uncertainty_notice: boolean, response_depth: score, }
A polished OpenAI + TypeSafe Jev terminal interface for answers with transparent decision reports
questions: { person_name: boolean, email_or_phone: boolean, postal_address: boolean, // +5 }
CLI that finds PII in text with TypeSafe Jev: presence, sensitivity, and located spans
questions: { ok: boolean, }
Unofficial CLI for the TypeSafe System One API (Jev) — designed for AI agents and anything that can execute a process.
questions: { urgent: boolean, severity: score, }
Node client for decision models such as Typesafe Jev
questions: { succeeded: boolean, }
CLI for TypeSafe AI's Jev evaluation model — typed questions in, structured JSON answers out
questions: { intent: choice, urgent: boolean, frustration: score, }
n8n community node for TypeSafe Jev structured AI decisions
questions: { q: boolean, }
jevkit for JavaScript/TypeScript: static linter and shared record format for building on TypeSafe's Jev (System One) model. Unofficial.
from typesafe_sdk import AsyncTypeSafeClient, Choice, Noul, Score
Build calibrated AI classifiers from human feedback using Jev and GEPA.
"name": "Jev",
Jev-style parallel constrained decisions for any MLX model on Apple Silicon. Typed, schema-valid JSON in one forward pass.
# baseUrl: https://api.typesafe.ai
A linter for the things a linter could never check: whether a function does
@app.post("/v1/systemone", response_model=DecisionResponse)Turn any off-the-shelf LLM into a Jev -like decision layer
The transport: one POST to `/v1/systemone` per call.
TypeSafe Jev for OTP: reply to Jev from a GenServer and pattern match on its answer
import { type Questions, TypeSafeClient } from "@typesafe-ai/sdk";Command-line tool for TypeSafe's Jev AI model
GROUPS = ('jev', 'llm')Jev vs GPT-6 Astra vs GPT-4.1 mini: direct Cartesian control of an xArm7 in MuJoCo.
/// Defaults to `https://api.typesafe.ai`. Path prefixes are preserved when the
Typed TypeSafe AI clients for Rust, with async and blocking backends and observable retries.
import { choice } from '@typesafe-ai/sdk';Interactive TypeScript coding CLI powered by Jev typed decisions and constrained AST generation.
from typesafe_sdk import ChoiceAnswer, JSONContent, NoulAnswer, ScoreAnswer, SystemOneResponse
Python 3.14. Get an API key from console.typesafe.ai and put it in .env:
DefaultBaseURL = "https://api.typesafe.ai"
Go SDK for the TypeSafe AI API — typed questions in, probability distributions out.
// SystemOne asks questions about a state via POST /v1/systemone.
unofficial go sdk for typesafe ai
* `POST /v1/systemone` の応答のうち、判定に必要な部分だけを検証する。
Zod validates the shape. Jev validates the meaning.
defaultAPIURL = "https://api.typesafe.ai/v1/systemone"
Semantic SQL for Postgres, powered by Jev
-X POST https://api.typesafe.ai/v1/systemone \
Agent-first Haskell DSL for TypeSafe's Jev judgment model: typed packets, inferred types, answers under the same labels
"typesafe": Backend("typesafe", "https://api.typesafe.ai/v1/systemone", "jev-latest", "TYPESAFE_API_KEY"),sort by meaning: order lines along a plain-English dimension, from pairwise comparisons judged by TypeSafe's Jev model
DEFAULT_URL = "https://api.typesafe.ai/v1/systemone"
Typed semantic decisions for Unix pipelines and CI, powered by TypeSafe AI Jev.
"endpoint": "https://api.typesafe.ai/v1/systemone",
Small dependency-free CLI for TypeSafe Jev
source: "jev" | "local";
Reusable machine-learning primitives for Jev — PCA, MCMC, text diffusion, neural cellular automata, and a task harness that picks the right tool.
baseURL: URL = URL(string: "https://api.typesafe.ai")!,
An independent, type-safe Swift SDK for TypeSafe Jev, with async/await, batching, retries, and SPM support.
/// Retrying is safe for this API: <c>POST /v1/systemone</c> is a stateless evaluation call, and a
Community .NET SDK for the TypeSafe AI System One API — typed noul, choice, and score questions with structured, confidence-scored answers. Not affiliated with TypeSafe AI.
val DefaultBaseUrl: Uri = Uri(scheme = "https", host = "api.typesafe.ai")
Scala SDK for Jev. No effect system bundled.
public static let defaultBaseURL = "https://api.typesafe.ai"
Unofficial Swift library for the TypeSafe API
var request = HttpRequest.newBuilder(URI.create("https://api.typesafe.ai/v1/systemone"))Jev from plain Java 25. No framework, no dependencies, one HTTP call.
SYSTEM_ONE_PATH = "/v1/systemone"
daf-jev: composable Python toolkit for TypeSafe's Jev (System One) decision API — question builders, confidence gates, evaluator, calibration, CLI, MCP server, agent skill
const response=await fetch('https://api.typesafe.ai/v1/systemone',{method:'POST',headers:{Authorization:`Bearer ${c.JEV_API_KEY}`,'Content-Type':'application/jsA small programming language for LLM workflows, powered by Jev.
--endpoint URL default https://api.typesafe.ai/v1/systemone
Ask Jev typed questions from the shell: noul, choice, and score answers as numbers, not prose
endpoint: "https://api.typesafe.ai/v1/systemone".to_string(),
High-throughput synthetic & pretraining dataset sifter powered by TypeSafe AI Jev (api.typesafe.ai). Stream, filter, and score Parquet & JSONL datasets at 1,500+ rows/sec using System One typed decisions (Choice, Score, Noul).
DefaultBaseURL = "https://api.typesafe.ai"
Go client for TypeSafe's System One API and its model Jev: typed judgments and calibrated probabilities instead of generated text
self.send(Method::POST, "/v1/systemone", Some(body), options, |raw| {Independent async and blocking Rust SDK for the TypeSafe AI System One API
public static let defaultBaseURL = URL(string: "https://api.typesafe.ai")!
Swift SDK for TypeSafe AI
* rooted at `https://api.typesafe.ai` and authed with the bearer token.
A Scala 3 / ZIO library for TypeSafe AI's
import { choice, score, TypeSafeClient } from "@typesafe-ai/sdk";A small local demo where a Mermaid flowchart contains the route rules, Jev supplies typed judgments, and JavaScript follows the matching branches.
import type { Fetch, Questions, SystemOneResult } from "@typesafe-ai/sdk";Semantic schemas over TypeSafe's Jev — validate the state locally, then project typed answers.
import { TypeSafeClient } from "@typesafe-ai/sdk"k8s awareness with jev
@default_base_url "https://api.typesafe.ai"
An idiomatic, type-safe Elixir port of the official TypeScript AI SDK (ai / ai-sdk) providing unified LLM integrations, streaming text and structured outputs, tool calling, and agentic workflows. Jev is their current flagship model and is the first System One model.
: (vercelGateway ? GatewayProtocol.DEFAULT_BASE_URL : "https://api.typesafe.ai");
Idiomatic Java SDK for TypeSafe AI Jev System One decision engine
import type { ChoiceResponse, EntryType, NoulResponse, Question, Usage } from "@typesafe-ai/sdk";Semantic test matchers for Vitest and Jest, powered by TypeSafe's Jev model. Write expectations in plain English, get calibrated probabilities back.
help = "API root [default: https://api.typesafe.ai]"
Interact with "Jev" model from TypeSafe AI
'description' => __( 'TypeSafe API base URL (default https://api.typesafe.ai/v1).', 'ai-provider-for-jev' ),
Connect WordPress to TypeSafe's Jev System One model for structured decisions (choice, score, noul).
defaultBaseURL = "https://api.typesafe.ai"
Community Go SDK for TypeSafe AI Jev / System One
import { choice, noul, score, TypeSafeClient } from "@typesafe-ai/sdk";TypeSafe AI'ın karar modeli Jev ile basit, tip güvenli bir örnek. Aynı görevi Claude Opus 5 ile yapan bir karşılaştırma scripti de içerir.
// request example for POST /v1/systemone (api.md:153), and its example
A Go client for the TypeSafe System One API — typed judgments and probabilities, zero dependencies outside the standard library.
self.url = "https://api.typesafe.ai/v1/systemone"
⚡ DOOM-JEV: Autonomous ViZDoom Agent
import { choice, noul, TypeSafeClient } from '@typesafe-ai/sdk';An autonomous agent that plays Clash Royale on macOS through the native
endpoint: "https://api.typesafe.ai/v1/systemone",
Fast Rust CLI for TypeSafe Jev: typed decisions, offline linting before you pay
'base_url' => env('TYPESAFE_BASE_URL', 'https://api.typesafe.ai'),Unofficial Laravel integration for TypeSafe Jev AI with typed responses, async requests, scoped dependency injection, and testing fakes.
API = "https://api.typesafe.ai/v1/systemone"
TypeSafe's Jev is a decision model: it takes state and typed questions and returns calibrated
/// Sends `state` and `questions` to `POST /v1/systemone` with client defaults. Returns decoded answers.
Rust SDK for the TypeSafe AI API
DefaultBaseURL = "https://api.typesafe.ai"
unofficial go SDK for typesafe AI, with typed answers, retries, and context support
* Typed question specs in the wire shape of `POST /v1/systemone` (PLAN.md §2.2).
Play Slay the Spire 2 with Jev (TypeSafe System One).
DefaultBaseURL = "https://api.typesafe.ai"
Go client for TypeSafe AI's System One API (Jev), with optional Langfuse instrumentation
"jev" => { input_per_mtok: 0.042, output_per_mtok: 0.0 },Community Rails integration for TypeSafe AI's System One API, built on the
/** TypeSafe's public /v1/systemone API. Thread safe; reuse a client across requests. */
Unofficial Java SDK for TypeSafe Jev and Vercel AI Gateway, with Spring Boot and WebClient support
export { choice, noul, score } from "@typesafe-ai/sdk";Typed, policy-driven decision workflows on top of TypeSafe AI Jev: confidence routing, fallbacks, evaluation, and RAG patterns for TypeScript apps.
_systemOneUri = _pipeline.Resolve("/v1/systemone");.NET SDK for the TypeSafe AI platform
//! `serde` shapes for `POST /v1/systemone` — no HTTP calls yet.
Jev-routed client for NEAR AI Cloud inference and IronClaw agents.
from typesafe_sdk import Choice, Noul, Question, Score
A thin shim between Pydantic and Jev.
import { APIError, AuthenticationError, TypeSafeClient, choice, noul, score } from "@typesafe-ai/sdk";Audit your git diff against YAML coding-standards packs using TypeSafe's Jev model, from a CLI or your AI agent's command/skill.
#define JEV_DEFAULT_URL "https://api.typesafe.ai/v1/systemone"
SQLite extension that calls TypeSafe Jev (or tensai serve) from SQL
public const DEFAULT_BASE_URL = 'https://api.typesafe.ai/v1';
PHP SDK for TypeSafe's Jev: send text and typed questions, get typed answers with calibrated confidence. PHP 8.1+, works with any PSR-18 client, Laravel 8–13.
import { choice, noul, type Questions, type SystemOneRequest, type SystemOneResult } from "@typesafe-ai/sdk";A step-through logic interpreter for natural-language facts and rules, unified with Jev
use jev_sdk::{Choice, Noul, NoulCriteria, Question, Score, TypeSafeClient};An agent-facing command line for Jev,
public static final String DEFAULT_BASE_URL = "https://api.typesafe.ai";
Community Java client for the TypeSafe System One API (unofficial)
import { experimental_evaluate as evaluate } from 'ai';Recursive Jev choice over a taxonomy. Select from more than 255 options without breaking TypeSafe Jev's choice cap.
SystemOneResult,
Run Laya, the open-source Jev-compatible System-1 decision model, from Node.js / TypeScript via ONNX Runtime
API_URL = "https://api.typesafe.ai/v1/systemone"
Fast CVSS scoring from vulnerability descriptions using Typesafe Jev
Real mode: uses the official ``typesafe_sdk`` (``pip install jev-triage[jev]``),
Triage for deep-research agents. Score N search results in one parallel TypeSafe Jev call, so the frontier model only reads what is worth reading.
endpoint: options.endpoint ?? "https://api.typesafe.ai/v1/systemone",
Typed, confidence-aware AI decisions for TypeScript and Python with Jev, OpenRouter, safe fallbacks, and per-call cost tracking.
from typesafe_sdk import Choice, Noul, NoulCriteria
atlas-jev is a local memory store. You give it text. An OpenRouter chat model extracts memories worth keeping, and those memories are stored on disk in LanceDB. You can search them
{ value: "jev-latest", label: "jev-latest" },shadcn-style reusable components and blocks for using TypeSafe AI.
export const SYSTEM_ONE_URL = "https://api.typesafe.ai/v1/systemone";
Verbatim context compaction for pi, powered by the TypeSafe Jev model.
pub const DEFAULT_BASE_URL: &str = "https://api.typesafe.ai";
Async-first Rust SDK for the TypeSafe AI API
DEFAULT_BASE_URL: Final = "https://api.typesafe.ai"
Async Python client for TypeSafe Jev. Typed questions in, probabilities and choices out, no prose to parse.
USING (VALUES ('API_URL', 'https://api.typesafe.ai/v1/systemone'),Ask your Db2 for i tables questions. A Db2 for i SQL SDK powered by TypeSafe's Jev.
DefaultBaseURL = "https://api.typesafe.ai"
Unofficial Go client for TypeSafe's System One API and its model, Jev.
/// Blocking `POST /v1/systemone`.
Rust client for the TypeSafe AI System One API: send a state plus named,
ENDPOINT = 'https://api.typesafe.ai/v1/systemone'
Working Memory Jev · Passage
'/v1/systemone',
PHP & Laravel SDK for TypeSafe AI's JEV Model series
"url": "https://api.typesafe.ai/v1/systemone",
这是在本机运行的 API 客户端,推理由云端 Jev 执行,不需要显卡。Jev 适合分类、判断和评分,返回选项、概率及分数;它不生成聊天回复或代码。
model: str = "~typesafe/jev-latest",
Directory Search with OpenRouter Decisions
@url = "https://api.typesafe.ai/v1/systemone"
Probabilistic control flow for Ruby — chance, pick, and rate, powered by TypeSafe's Jev
DefaultBaseURL = "https://api.typesafe.ai"
Go SDK for TypeSafe AI.
printf ' 官方直连 key(api.typesafe.ai)。留空跳过,之后可 export TYPESAFE_API_KEY。\n'
A CLI for the TypeSafe Jev System One decision model — defaulting to the TypeSafe direct API (api.typesafe.ai), with OpenRouter Decisions as a switchable alternative — plus a pi ex
if want := "https://api.typesafe.ai/v1/systemone"; call.URL != want {Go SDK for TypeSafe AI — classification and rating primitives over text and JSON
const DefaultBaseURL = "https://api.typesafe.ai/v1"
Unofficial Go SDK for TypeSafe AI's Jev API
.post(Uri.parse('https://api.typesafe.ai/v1/systemone'),Run Flutter integration tests using natural language with TypeSafe.ai's Jev
pub const DEFAULT_BASE_URL: &str = "https://api.typesafe.ai";
Unofficial typed async Rust client for the TypeSafe System One API
public static final String DEFAULT_BASE_URL = "https://api.typesafe.ai";
A Java client for TypeSafe's System One API and its Jev models: send some state and a set of
.post("https://api.typesafe.ai/v1/systemone")DAG creation out of unordered items via Jev
constructor(apiKey: string, baseUrl = "https://api.typesafe.ai/v1") {Zero-dependency TypeScript client for TypeSafe Jev (System One decision model) — OpenRouter, TypeSafe direct, and Vercel AI Gateway providers
POST /v1/systemone Jev-compatible (TypeSafe's shape): {state, questions, model}convaiinnovations/laya 的本地推理引擎:不训练、不微调,把官方发布的三份权重跑成一个 HTTP 决策服务,Mac(Apple 芯片)上开箱可跑;同时复刻 TypeSafe Jev 的请求/响应格式,按官方接口写的客户端只改一个 base URL 就能打到本地权重,其余代码、类型、校验都不用动。
api_base: str = "https://api.typesafe.ai/v1",
Turns Jev into a chatbot
import { TypeSafeClient } from '@typesafe-ai/sdk';Standalone TypeSafe Jev code-review CLI: typed decisions over a local git diff.
#define JEV_DEFAULT_URL "https://api.typesafe.ai/v1/systemone"
Batched natural-language judgments for SQLite, powered by TypeSafe Jev
from jev_engine import SystemOneClient, Noul, Choice, Score, SystemOneResult
Jev-Engine: 面向软件工程的 System One 离散决策引擎与完整上下文投影切片管道
| `:base_url` | `TYPESAFE_BASE_URL` | `https://api.typesafe.ai` |
Unofficial Elixir client for TypeSafe's System One API (Jev): typed Noul, Choice and Score judgments over Req
pub const DEFAULT_BASE_URL: &str = "https://api.typesafe.ai";
Unofficial Rust client for TypeSafe's System One API (Jev): typed Noul, Choice and Score questions, async and blocking
TypeSafe's docs warn that ``jev-latest`` moves and that tuned confidence
jevkit for Python: static linter and shared record format for building on TypeSafe's Jev (System One) model. Unofficial.
url = "https://api.typesafe.ai/v1/systemone",
A native R client for Jev System 1 model decisions
const kDefaultBaseUrl = 'https://api.typesafe.ai';
Typesafe Jev Api
e = estimate({"state": "...", "model": "jev-latest", "questions": {...}})Unofficial CLI + Python estimator of tokens, cost and context limits for TypeSafe (System One / Jev) API requests. Not affiliated with TypeSafe.
import type { EntryType } from '@typesafe-ai/sdk';Rule-based code review for AI coding agents, powered by TypeSafe Jev.
import { experimental_evaluate as evaluate } from 'ai';A vocabulary-driven TypeScript runtime for safe, stateful applications powered by TypeSafe AI Jev.
data, response, err := c.execute(ctx, http.MethodPost, "/v1/systemone", body, resolved)
A Go SDK for the TypeSafe API, enabling quick integration with Jev.
default: 'https://api.typesafe.ai',
Helper n8n community node for Jev by TypeSafe. Classify, route, and score text with questions you define, and get a probability for every answer so unsure items can go to review.
* Non-blocking client for TypeSafe Jev's {@code /v1/systemone} endpoint.Non-blocking Spring Boot Starter for Spring WebFlux that implements a Semantic Circuit Breaker to detect silent HTTP 200 failures using TypeSafe Jev.
API_URL = "https://api.typesafe.ai/v1/systemone"
Local console + CLI for TypeSafe's Jev (System One) model: send a state and typed questions, get calibrated probabilities back.
DefaultBaseURL = "https://api.typesafe.ai"
Typed answers from Jev, in idiomatic Go.
endpoint: "https://api.typesafe.ai/v1/systemone".to_string(),
System One TypeSafe AI Rust Client (unofficial)
defaultBaseURL = "https://api.typesafe.ai"
Idiomatic Go SDK for the TypeSafe AI API