A curated directory of Jev integrations, compatible models, and System One tools.

Jev Ecosystem

Every project here is active and in use: pushed to within the last two weeks, 50+ stars, not archived, with a clear description of what it does with Jev. No weekend scaffolds, no dead forks.

Unofficial and community-maintained. Not affiliated with or endorsed by TypeSafe AI.

What is Jev?

Jev is the first of TypeSafe AI's System One models: a model that takes unstructured state plus a typed question and returns a typed decision — a choice, a score, or a boolean — each with a calibrated confidence. It is not a chat model. It is a decision function you call from software, which makes it a natural fit for classification, routing, verification, ranking and agent guardrails.

	Chat LLM	Jev
Input	Prompt	State + typed question
Output	Free text to parse	Typed answer with confidence
Used for	Generation	Decisions

_{Price per million tokens. Source: TypeSafe AI launch post, September 2026.}

How entries are chosen

An entry qualifies when all of the following hold:

1. Alive — a push within the last 14 days at the time of review
2. Used — 50 or more GitHub stars
3. Real — describes a concrete typed-decision use of Jev (or a documented Jev-like model); not a generic classifier that merely resembles the pattern
4. Open — public repository, not archived, not a fork

Official TypeSafe AI repositories are listed regardless of stars. Entries marked integration are projects that existed before Jev and added it; the rest were built for it.

Listing is not endorsement. Entries are collected from public sources and filtered by activity and usage. We check that a project is active and describes a real Jev use; we do not audit its code or verify its published numbers.

Contents

• Official — 5
• Models — 34
• SDKs & Clients — 3
• Frameworks & Integrations — 31
• Agent Tooling — 50
• Browser & Computer Use — 18
• Classification & Routing — 9
• Verification & Guardrails — 10
• Scoring & Ranking — 11
• Applications — 24
• Games & Simulations — 7
• Benchmarks & Research — 18
• Demos & Playgrounds — 1

Official

Repositories published by TypeSafe AI itself.

• TypeSafe agent skills ★ 2.6k - installable agent skills package (npx skills add typesafe-ai/skills) that teaches agents the Jev workflow
• System One adapter (Python) ★ 383 - Official drop-in TypeSafeClient replacement backed by OpenAI, Anthropic, and compatible LLM APIs, for comparing Jev against chat models
• TypeSafe Python SDK ★ 276 - The official Python library for the TypeSafe API
• TypeSafe JavaScript SDK ★ 270 - Official TypeScript/JavaScript client with inferred answer types. npm install @typesafe-ai/sdk
• TypeSafe Daggerverse ★ 24 - Official collection of reusable Dagger modules for TypeSafe AI and System One workflows

Models

Open-weight and open-source models that reproduce or extend the System One / Jev approach: typed questions in, calibrated typed decisions out.

• laya ★ 31.7k - Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request
• kev ★ 8.7k - Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
• CLM ★ 2.9k - Contrastive language model for one-pass typed decisions and option probabilities
• deepopen ★ 2.1k - System 1 DeepOpen Multilingual, non-autoregressive System 1 decision engine
• NeoHorse-Jev ★ 1.7k integration - Open 4B decision model with prefill-only Choice, Noul, and Score inference over text or one image, plus native, vLLM, and SGLang runtimes
• Jeff ★ 1.5k - Millisecond decisions, any domain: a 0.8B open "System 1" model that picks between your options with calibrated probabilities. One base, swappable LoRA adapters, on your own hardware
• ollaya ★ 1.3k - Run open decision models locally: pull and serve Laya, decider, NLI and GLiClass behind a TypeSafe-compatible API. Ollama for decision models
• jevos ★ 1.2k integration - jevos is an open-source alternative to Jev for yes/no decisions that runs on your laptop
• AnyJev ★ 1.1k - Turn any LLM into a Jev-style decision model: typed decisions, real probabilities, no training. (continue updating, welcome any issue and PR request)
• TypeLLM ★ 937 - Type-safe language-model generation and structured decisions
• von ★ 875 - The open-source System One decision model. Sub-15ms, non-autoregressive, local drop-in alternative to TypeSafe Jev
• rizzo-flow ★ 867 - The open, local take on Jev: typed decisions from an LLM, without generating a single token
• Valen ★ 699 - Train a Jev-like multimodal model by yourself. System One Model, now with vision
• openjev ★ 657 - Open, Jev-compatible System One decision server on DiffusionGemma
• Intern-Decision ★ 654 - Fast multi-modal decision model
• Simple Jev ★ 595 - Turn any open model into a classifier/jev endpoint
• Jeeves ★ 415 - Jeeves – Reasoning improves Jev-like decision models
• LLM2Jev ★ 411 - Turn local language models into Jev-style structured decision models. Get results from text and images with prefill alone—no token-by-token decoding required
• Open-Jev ★ 397 - Open probability decisions with Qwen3.5-2B, Qwen3.5-9B and Qwen3.8-27B
• agent-jev ★ 374 - AgentJev-0.6B - a fast 'System One' decision model for AI Agents: feed it any unstructured state (diffs, traces, logs) and structured questions, get calibrated probability distributions back in one ~50ms forward pass.…
• imajev ★ 344 - Open Jev-style typed-decision model that also takes images: photo + app state + typed questions in, calibrated probabilities out, locally
• tev1 ★ 226 - Open-weight, Jev-inspired decision model finetuned on top of Qwen3.5 4B
• djev-spark ★ 222 - DiffusionGemma NVFP4 structured decisions on a DGX Spark: container recipe
• OneJev ★ 147 - 🚀🚀 A multimodal System One decision model that gives calibrated answers to typed questions about screens, photos, video and text in one forward pass
• JevK5 ★ 144 - JevK5: open-weight alternative to TypeSafe Jev. Typed decisions with probabilities in one forward pass; Apache-2.0 weights and code
• djev ★ 113 - Jev-style structured decisions on DiffusionGemma: the example server from vLLM PR 57250
• laya-server ★ 96 - A self-hosted API and web interface for Laya’s structured decision models, compatible with the TypeSafe Jev API format
• open-jev-fast ★ 90 - Faster inference backend for Open-Jev-27B: fused CUDA kernels, prefix tree, CUDA Graphs (B300, bf16)
• Phocinae-Largha-150M-v1 ★ 75 - Bilingual (EN/ZH) 144.3M open decision model: noul / choice / score answers with calibrated probabilities, served locally behind a TypeSafe-compatible SystemOne API
• JevAny ★ 70 - Open infrastructure for training, evaluating, and deploying System 1 decision models across language and multimodal backbones
• stuntd ★ 69 - Local proxy that learns your app's typed LLM decisions and answers them with a Laya head. Jev and OpenAI compatible
• JevEmbed ★ 66 - Meet JevEmbed — turn embeddings into decisions. Choose, score, and judge with your choice of embedding model
• OpenDecision ★ 58 - OpenDecision is an open-source semantic decision engine like typesafe's jev
• jevify ★ 54 - Supersimple way to serve LLMs as a Jev-like endpoint
• sys1 ★ 51 - Blazing fast, self-hosted structured decisions for open models with a TypeSafe AI compatible API, written in Rust

SDKs & Clients

Client libraries for calling Jev from other languages and runtimes.

• openai-scala-client ★ 255 integration - Scala client module for TypeSafe System One that submits Jev Noul, Choice, and Score questions and returns typed answers
• ruby_decision_model ★ 56 - Ruby client for decision models such as Typesafe Jev
• spring-ai-typesafe ★ 52 - A Java SDK for the TypeSafe AI JEV API, & Spring AI TypeSafe integrations

Frameworks & Integrations

Jev wired into existing frameworks, gateways, databases and developer tools.

• litellm ★ 60.4k integration - AI gateway with a TypeSafe guardrail and Jev classifier for policy judgments and complexity-based request routing
• oh-my-pi ★ 34.7k integration - Coding agent with a TypeSafe judgment provider that uses Jev for typed checks, routing, and workflow decisions
• Sim ★ 29.8k integration - open-source collaborative workspace for building, deploying, and monitoring AI agents featuring native TypeSafe System One evaluation and decision provider integration
• Pydantic AI ★ 20.5k integration - official Pydantic AI agent framework shipping first-class TypeSafeModel integration to map Pydantic schema fields into typed Jev System One questions with confidence scoring
• json-render ★ 18.6k integration - Generative UI framework with an experimental evaluator that lets Jev choose among application-supplied composition criteria
• laya-mlx ★ 6.8k - Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API
• eve ★ 5.5k integration - Vercel's open agent framework, which ships Jev as the default evaluation model in its experimental evaluate path
• LangWatch ★ 4.9k integration - LLM evaluation platform whose Instant Evals sends production traces to Jev for batched Noul, Score, and Choice judgments
• RubyLLM ★ 4.4k integration - official Ruby gem connecting TypeSafe judgment models to RubyLLM with a native System One protocol for typed questions, probabilistic answers, and error normalization
• Milvus Search with Jev ★ 2.4k integration - nine runnable notebooks combine Gemini embeddings and Milvus retrieval with Jev Noul and Choice judgments, while Python applies ranking, filtering, routing, and stopping policies to synthetic examples
• ai ★ 1.2k integration - Laravel AI SDK with a TypeSafe provider for Jev Choice, Score, and Noul classification requests
• pg-jev ★ 1.1k - Ask your Postgres tables questions in plain language. A PostgreSQL extension powered by TypeSafe's Jev
• ai-cli ★ 823 integration - Vercel Labs terminal CLI that can run Jev as the evaluation model for its evaluate command
• aiavatarkit ★ 689 integration - Conversational-avatar framework with a Jev gate that judges whether a speaker has finished a turn
• djev-run ★ 580 - deploys DiffusionGemma-Jev behind a TypeSafe-compatible API on a Cloud Run GPU with snake, dino and tetris demos wired to the decision endpoint
• smithers ★ 430 integration - Agentic TypeScript workflow framework with a Jev session checker wired into its workflows
• vLLM Jev ★ 359 - Native vLLM serving for Jev decision models
• JEV Sees ★ 306 - Eyes are All JEV Needs - real time visual devisions from RGB, video and RGB-D cameras
• JarvisCore ★ 242 integration - Python multi-agent runtime that ships Jev natively from 1.12, where agents ask typed Choice, Score and Noul questions through a decision client separate from the text model, the Kernel picks a specialist subagent…
• ai-python ★ 192 integration - the official Vercel AI SDK for Python carries Jev through its evaluation operation and Gateway examples
• jev4pg ★ 175 - Established strong JEV based natural language to SQL performance and well-covered semantic operators for PostgreSQL(PG) with optimized cost and latency. Reusable evidence, with native execution, parallel plans and…
• holon ★ 153 integration - keeps decisions in their own crate and notes in the adapter for its first backend, Cloudflare's Clef, that "Clef uses the System One-shaped state/questions request" — the interface being borrowed rather than reinvented
• NeuroLink ★ 144 integration - The pipe layer of an AI nervous system — one interface connecting provider neurons to your application, across three inference types: generate, stream, and a calibrated decide (via TypeSafe Jev). MCP-native, voice…
• jevals ★ 102 - Agent evals and guardrails as Jev decisions: one request per trace, a fraction of a cent, fast enough for the agent loop. Runs locally with Kev or Laya
• pg_typesafe ★ 88 - Pre-alpha PostgreSQL extension for TypeSafe AI (Jev) categorical classification
• laya-php ★ 87 - Classify text in PHP without an LLM bill: typed decisions in 100+ languages, self-hosted. Laravel-ready SDK for Laya, a Jev AI alternative
• system-one ★ 70 - Provider-neutral System One runtime for TypeScript and Pi
• jev-foundation-models ★ 66 - A lightweight, native Swift 6 bridge integrating TypeSafe AI's Jev System One decision model into Apple's Foundation Models framework
• beam_weaver ★ 64 integration - BeamWeaver exposes Jev Choice, Score, and Noul evaluations as provider-neutral decision-model calls in Elixir workflows
• grev ★ 55 - grep, sort, cut and uniq that match by meaning — grev 'is a vegan meal' menu.txt keeps the lines whose Jev Noul clears 0.5, sibling filters route by Choice and rank by Score, and the output is always your own…
• laya-jev-GraphRAG ★ 50 - A database-agnostic Agentic GraphRAG framework using swappable System One models (local Laya / cloud Jev). A plug-and-play intelligence layer featuring a complete 4-phase pipeline, continuous evaluation and custom A*…

Agent Tooling

Jev as the decision layer inside coding agents and agent frameworks: model routing, permission gates, compaction, planning.

• AutoGPT ★ 187.5k integration - open-source autonomous agent platform featuring first-class TypeSafe Jev decision blocks for typed routing, filtering, scoring, and confidence-gated next-action dispatching
• oh-my-claudecode ★ 39.7k integration - multi-agent team orchestration for Claude Code featuring opt-in Jev hooks for sub-millisecond judgment points, decision caching, and per-point egress controls
• jcode ★ 20.4k integration - RAM-efficient autonomous agent harness implemented in Rust with native TypeSafe Jev typed decision transport for memory pruning, browser navigation, and voice interaction routing
• Eliza ★ 19.6k integration - multi-agent framework integrating TypeSafe System One decision services for sub-100ms intent classification, action dispatching, and confidence-gated tool execution
• jev-chat-jarvis ★ 7.5k - The chat decision assistant: before you reply, Jev reads the chat, judges intent and risk, and drafts replies you fill in with one tap. You press send. Android · Windows · macOS · iOS
• jevgrep ★ 2.5k - Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context
• agentconnect ★ 1.5k integration - Team agent workspace using Jev typed decisions for response routing and model selection
• hermes-jev-skills ★ 1.1k - Jev-powered model routing, memory, compaction, skill selection, computer and browser use for Hermes agents (also Claude Code and Codex)
• Atomic ★ 850 integration - ships a first-class Jev structured-output provider so an agent's decisions come back typed, through the same decision resolver as its other providers
• JevRev ★ 706 - An LLM + Jev workflow that changes EVERYTHING. Boost your vertebrate brain with a spine inside
• distill ★ 692 integration - Coding-agent runtime using Jev decisions for model routing, context retention, and tool selection
• jev-mcp (jkudish) ★ 510 - Fast, cheap, typed judgments from TypeSafe's Jev model, as MCP tools
• mu ★ 456 - Coding-agent workspace where Jev or a local Laya judge gates bounded decisions such as context retention, approvals, routing, drift, and inter-agent sharing
• vexjoy-agent ★ 439 integration - VexJoy AI Agent with Jev Intelligent Routing - /do routes plain-English requests to the right specialist agent and gates the work with reviews, tests, and a learning loop
• typesafe-mcp ★ 346 - System One MCP connector to evaluate anything fast and cheap. Give your AI agent direct access to models like: Jev, D1, CLM and Laya
• Keel ★ 330 - Local-first macOS coding workspace with local Laya and optional Jev decision selection
• jev-gateway ★ 308 - An easy way to use jev with your coding agent for tool calling reasoning
• Astra-Ares ★ 303 - Adaptive reasoning effort for GPT-6 during Codex tasks, powered by Jev to reduce token usage
• decapod ★ 236 integration - During assurance evaluation, Jev estimates how likely a proposed agent trajectory is to satisfy its bounded objective
• Albatross ★ 235 integration - Terminal-first coding agent with an experimental Jev Mode that puts the decision model in front of the main LLM to answer supported error questions directly
• compact-adviser ★ 206 - Agent plugin that uses Jev to judge when a coding session should compact its context
• Jev-Mem ★ 190 - Jev-Mem: System-One Controlled Agentic Memory
• deepseek-harness-jev ★ 189 - Native DeepSeek Harness (DSH) plugin integrating TypeSafe Jev as a System One decision layer for agent selection, supervision, corrections, and approvals
• glowbom-oss ★ 164 integration - Coding agents can ask Jev to judge small workflow decisions, such as whether build results look healthy
• pi-warden ★ 162 - Guardrails for Pi built on pi-typesafe that steer the agent instead of interrupting you: Jev judges irreversible and off-task tool calls, detects stuck loops, checks unverified done claims, flags slop
• jev-pruner ★ 159 - Claude Code plugin: trim long Bash output with TypeSafe Jev before the model sees it
• Anna-Agent ★ 147 integration - Anna uses Jev as a fast judgment layer for bounded enterprise-agent recommendations and abstentions
• jgrep ★ 136 - grep, but the pattern is a description. Filters lines by meaning with TypeSafe's Jev decision model: ~200 ms and a thousandth of a cent per line
• Building with TypeSafe Jev ★ 136 - Unofficial skill that teaches coding agents to build with TypeSafe AI's Jev: typed decisions, calibrated confidence, and prior art from 150+ community projects
• skillranker ★ 129 - Rust CLI powered by Jev from TypeSafe.ai that ranks agent skills for the next step using live session context. Includes Claude Code hooks, structured JSON, abstention, and local feedback. Requires a TypeSafe API key
• jevmem ★ 122 - jevmem saves the decisions, rules and failed approaches from your Claude Code chats to JEVMEM.md in your repo, and brings the relevant ones back next session
• jev-lint ★ 119 - lint text in code by jev scorerer
• Quicksilver ★ 111 - Claude Code skill: hand bulk judgment calls to Jev. 86% fewer Claude tokens on a 12-task benchmark, up to 20x faster. One-line npx install
• agent-router ★ 111 - CLI that picks Cursor, Claude Code, Codex, or OpenCode + model/effort for a task, then launches it. Powered by Jev and Herdr
• jegrep ★ 110 - Semantic grep: find code by describing what you're looking for, powered by Jev
• skills ★ 106 integration - Four agent skills use Jev Score and Noul judgments to assess subjective criteria with calibrated probabilities
• winnow ★ 106 - A calibrated context sieve for Claude Code: every tool result is judged by a System One model before it enters context
• jevcore ★ 106 - TypeSafe Jev for DeepSeek Harness, the Model Context Protocol, and plain Node: typed judgments instead of prose, offline by default
• jevgrep ★ 101 - Jev-powered semantic code search for coding agents — find behavior across repositories via CLI or MCP, with exact source excerpts and line numbers
• grok-bot-jev ★ 99 - Connect TypeSafe Jev to Grok Bot as a cheap decision layer - usage gates, skill template, examples
• oxlint-plugin-jev ★ 85 - Oxlint rules written in plain English: each match is sent to Jev as a yes/no question and reported when the yes-probability clears a cutoff
• weflow-cli ★ 81 integration - When enabled, Jev classifies article topics and relevance for WeChat daily digests
• save-token-jev-clean ★ 80 - Portable, Jev-guided context compaction for coding agents
• captaincore ★ 75 integration - Jev triages malware-scan findings and supplies typed Noul, Choice, and Score decisions for WordPress maintenance
• total-agent-memory ★ 72 integration - Jev checks retrieved memories for contradictions before an agent uses them in an answer
• jev-judge-mcp ★ 71 - Typed judgment tools for MCP agents. TypeSafe's Jev model as verify, screen, find, classify, rerank, decide, compare, extract, review, gate, and score: the model judges, policy decides auto, review, or escalate
• jev-rules ★ 66 - Jev picks which of your rules apply to each prompt, so Claude only sees the ones that matter
• deepreseach-learing ★ 64 - LangGraph deep-research agent whose Jev shadow mode attaches a quality assessment to every generated report
• pi-jev (TheoOliveira) ★ 63 - Semantic tool routing and typed System One decisions for the Pi coding agent using TypeSafe Jev
• jmeter-ai ★ 63 integration - Jev classifies JMeter agent intent and selects a focused tool pack, with a full-tool fallback when routing is uncertain

Browser & Computer Use

Agents that drive browsers and desktops with Jev deciding the next action.

• cua ★ 29.0k integration - Computer-use platform with a Jev recipe where the model selects one bounded action candidate and the client validates it before execution
• Jev Ultrafast ★ 22.4k - Browser Use's ultrafast agent: Jev picks the operation and DOM element in one request; a small LLM only writes text when typing is needed
• agent-desktop ★ 1.8k integration - Desktop automation runtime with a Jev skill that chooses UI operations and targets from accessibility-tree candidates
• typesafe-computer-use ★ 1.2k - Computer use for about $0.0002 a step: OCR the screen, classify the next action with TypeSafe, click. macOS
• reticle ★ 1.2k integration - AI agents can generate code, but still struggle to understand what they build. Reticle brings Jev-style machine-native runtime perception to web & desktop applications
• celesto ★ 1.0k integration - Agent-computer platform with a runnable pull-request review lab where Jev judges whether findings are introduced, supported, and actionable
• surf-cli ★ 632 integration - Optional Jev commands find, filter, and verify browser elements, then select bounded actions subject to confidence and write-authority checks
• jev-browser ★ 319 - Browser use using Typesafe's Jev model
• arc-cua ★ 286 - Jev selects the next UI operation from the targets and actions exposed by the current desktop state
• FluidUse ★ 251 - Local computer use on Apple silicon: on-device typed decisions with laya and form filling with CUA-S1-FORMS, driven through the macOS Accessibility API
• jev-social ★ 158 - Open-source, local-first social media research agent for Instagram, TikTok, and LinkedIn. Jev routes read-only steps; socai CLI captures cited browser evidence
• system1-agents ★ 124 - System 1 decision models (Jev, Laya, Cua-S1) as brain for agents: Browser use, computer use, games and robotics
• Jevry ★ 123 - Desktop browser where Jev selects executable browser actions and a separate text model handles planning
• fastbrowse ★ 114 - A fast browser agent: Jev picks each action from what is on the page, an LLM reads and plans, and every claim in an answer cites a quote from the page
• jev-browser ★ 105 - Jev Browser — indexed browser automation. Jev chooses the control, Playwright acts. A CodexQA skill
• jev-browser (Ying-Kai-Liao) ★ 103 - Browser automation where an LLM plans and Jev (Typesafe System One) decides. Library, CLI and MCP server
• playwright-recast ★ 64 integration - Jev chooses camera moves, emphasis, and timing from a rendered Playwright recording and bounded direction candidates
• ghosthands ★ 56 integration - USB-HID GUI automation agent with an optional Jev fast lane that picks the next action, with a verified-DONE check before it stops

Classification & Routing

Jev sorting incoming state into categories or picking a destination: tickets, intents, documents, requests.

• SiYuan ★ 46.7k integration - privacy-first personal knowledge management system featuring native Jev decision model integration for high-speed document classification, flashcard intent categorization, and automated tag routing
• AgentScope ★ 32.9k integration - multi-agent platform by Alibaba implementing native TypeSafe Jev classification models for binary, choice, and score routing across agent pipelines
• inbox-zero ★ 12.4k integration - open-source AI email assistant that uses TypeSafe Jev System One decision models to classify incoming email intent and triage action items
• Paca ★ 1.9k integration - Project management app using Jev for agent routing, task fields, assignee selection, and automation conditions
• docjev ★ 520 - A very fast document classifier/splitter using Jev
• JevRouter ★ 449 - A lightweight Jev-powered router for models, tools, and subagents
• notra ★ 256 integration - Marketing analytics platform whose feature flag routes brand-visibility classifiers off an LLM and onto Jev boolean decisions
• GoEventBus ★ 73 integration - GoEventBus — high-performance event bus with a decision layer
• HA-Jev ★ 72 - Ask your house a question, get a number back. Home Assistant integration for TypeSafe Jev: typed answers as sensors, four actions for automations, and a conversation agent for Assist

Verification & Guardrails

Jev as a yes/no gate: verifying outputs, moderating content, enforcing policy.

• claude-code-templates ★ 32.5k integration - Claude Code configuration toolkit shipping Jev-based model routing, skill suggestion, and tool guardrail modules
• Dub ★ 24.9k integration - calls typesafe-ai/jev in malicious-link-check.ts before a short link is created, so the URL is gated by a typed verdict rather than a blocklist
• openwork ★ 24.0k integration - Open-source Claude Cowork alternative whose eval testkit can use Jev as a typed verification judge for agent-produced work
• fx ★ 3.3k integration - ships a typesafe_permission_reviewer builtin so the agent's permission decisions run through Jev rather than an LLM call
• Skill Scanner ★ 2.6k integration - Cisco's scanner hunts prompt injection and exfiltration in agent skills, and ships a System One analyzer as a deliberately advisory tier that cannot emit a finding or change a severity
• foreman ★ 717 - An agent supervisor and software factory foreman powered by TypeSafe’s Jev model
• abide ★ 565 - Supervises coding-agent edits and uses Jev to flag violations of project rules
• perch ★ 348 - Semantic code linting with Decision Models
• pi-jev (y0usaf) ★ 161 - TypeSafe Jev as a decision layer for the Pi coding agent: a measured tool-call gate plus jev_ask for typed, calibrated answers
• jev-guard ★ 63 - Auto mode for every coding agent, built on Jev: risk-scores every tool call with session context (deny / ask / allow), flags prompt injection in results, checks skills and plugins. Claude Code, Codex, Copilot, Gemini,…

Scoring & Ranking

Jev producing scores and orderings: rubric grading, reranking, labeling.

• WorldMonitor ★ 88.1k integration - real-time global intelligence dashboard using TypeSafe Jev questions to score news headline severity into 5 threat levels and categorize events across 14 conflict, cyber, and infrastructure domains
• OpenViking ★ 39.4k integration - Volcengine's agent context database ships a Jev rerank client that scores each candidate document with jev-latest against api.typesafe.ai and treats the returned probability as relevance, because TypeSafe exposes no…
• memsearch ★ 2.7k integration - Agent memory search with optional Jev relevance judgments to rerank retrieved passages
• hippo-memory ★ 774 integration - Agent memory library with an optional Jev reranker that judges which retrieved memories are relevant
• jev-search ★ 523 - Search the web with TypeSafe's Jev: source selection, query understanding and relevance ranking. Built with Search1API
• Vector Graph RAG ★ 254 integration - uses Jev Noul judgments to score candidate relations and applies a configurable threshold before retrieving their linked documents
• supercov ★ 151 integration - Code-quality and coverage reviewer that uses Jev to judge coding-agent changes
• jev-semgrep ★ 149 - grep by meaning, across languages. TypeSafe Jev scores every line against a meaning; combine meanings with AND/OR/NOT. grep
• Jev-Code-Reviewer ★ 140 - Review behavior, not just diffs. Jev prioritizes human attention; OpenAI explains the changes. Local CLI + agent skill + GitHub extension
• jev-seo ★ 97 - Free Rust SEO and GEO toolkit powered by TypeSafe Jev: 58-rule audits, site crawls, AI citation checks, rank drift, CI gates, 15-tool MCP. Zero cost
• jgrep ★ 59 - grep for what code does, not what it's called. Semantic code search powered by TypeSafe Jev

Applications

End-user products and vertical tools using Jev decisions.

• TradingAgents ★ 110.3k integration - Trading research agents with optional Jev screening of social posts for company relevance and sentiment
• ai-hedge-fund ★ 63.9k integration - multi-agent AI hedge fund trading system featuring native JevLLM integration to execute fast typed decisions without parsing fragility
• QuantDinger ★ 12.6k integration - Open-source AI Trading OS, agent trading, and vibe trading, with Jev System One integration. Research, build Python strategies, backtest, and paper/live trade across crypto, stocks, and forex. Launch your own…
• genoffice ★ 9.0k integration - Jev assigns a calibrated relevance score to each of the top 20 locally recalled files and reranks the search results
• xerj ★ 3.4k integration - Jev scores whether each retrieved document answers the query and reranks the search results by that relevance
• PyGPT ★ 2.0k integration - Desktop AI assistant with a Jev plugin for typed classification, routing, verification, and scoring
• jev-chat-windows ★ 767 - Windows WeChat reply assistant that uses local OCR and Jev to classify intent before presenting three manual-send replies
• tax-doc-classifier ★ 501 - Tax document page classifier built on Jev decisions. 100% strict accuracy across 261 IRS forms, ~$0.001 per page
• jev-chat-jarvis-mac ★ 471 - Jev classifies incoming WeChat or QQ messages by intent and risk before the app presents reply guidance
• horosa-skill ★ 450 integration - When explicitly enabled, Jev selects an astrology technique when keyword routing has no match and extracts settings stated by the user
• claude-code-trace ★ 377 integration - Jev scores agent progress, tool use, focus, exploration, recovery, and token efficiency from Claude Code session logs
• Goutoujunshi Jev Chat ★ 358 - Cross-platform chat assistant where Jev chooses a bounded reply strategy from reviewed conversation context before the user sends any draft
• AIStock ★ 347 integration - Stock-analysis application that uses Jev for structured market judgments
• crush-monitor ★ 281 - Crush Favorability Monitor: Use Jev to analyze the emotions, intentions and reply performance of WeChat chats. Deploy locally, use your own API Key
• summarize ★ 226 integration - The optional Jev prefilter scores transcript segments before the selected material is sent to the summarizer
• Jev-X-Sentiment-Analysis ★ 192 - An on-demand crypto market intelligence and decision-support terminal powered by TypeSafe AI's System One model (Jev)
• OmniStudio ★ 192 integration - OmniStudio exposes Jev Noul, Choice, and Score decisions for triage, scoring, content guards, retrieval reranking, and intent routing
• changelog.earth ★ 165 - When enabled, Jev performs bounded editorial and title checks on stories selected for an Earth changelog edition
• prism-liquidity-agent ★ 123 integration - Solana liquidity agent with optional Jev shadow judgments for deposit choice, toxic flow, and market-regime stress
• superterminal ★ 104 integration - Jev ranks command-palette entries against a plain-language query so the intended terminal action appears first
• formanator ★ 102 integration - Submit Forma https://joinforma.com benefit claims from the command line and Model Context Protocol (MCP) clients, with support for AI-powered receipt analysis with an LLM or Jev
• rikkahub-sillytavern-android ★ 89 - AI Jev RP QQbot AI Star Android AI Chatbox: lossless SillyTavern card import, aggressive cache optimization, Jev decisions, multi-dimensional memory, anti-blank-reply, proactive messages, AI group chat, SillyTavern…
• QuantStudio ★ 82 integration - Jev evaluates market state from completed candles, live prices, positions, and strategy constraints, then suggests observe, open, or close actions for human approval
• JevIntent ★ 66 - WeChat plugin that uses Jev to analyze message intent, emotion, and reply posture locally

Games & Simulations

Jev-driven bots, NPCs and simulated agents.

• jev-drone ★ 261 - Camera-only autonomous drone in MuJoCo with a small judgment model (TypeSafe Jev) in the loop at 2.5Hz
• beebots ★ 261 - Three AI trading bees on OKX, every decision by Jev. Paper trading by default. Not financial advice
• quackd ★ 255 integration - One CLI for all your robots. Connect them, command them, and let them work together, each with an LLM for a brain (Claude, OpenAI, Gemini, Grok, or local via Ollama or vLLM), VLAs for arms and decision LLMs (Jev, Laya,…
• Jev Plays Pokémon Red ★ 124 - Jev, an AI decision model, plays Pokémon Red. It beat the game in 37h 40m
• detroit-ai-player ★ 64 integration - Jev uses one typed Choice question to select each narrative action from the options available at a Detroit: Become Human decision node
• RoboJEV ★ 56 - Two-stage JEV control of a Franka Panda in MuJoCo
• PlayJev ★ 52 - 🚀🚀 A 0.8B JEV-like multimodal model playing GUI games directly from raw pixels

Benchmarks & Research

Evaluations, calibration studies and reproductions around System One models.

• Bespoke Nimble ★ 2.1k - Local typed decisions, contrastive data curation, and model evaluation
• minutes ★ 1.5k integration - local-first transcription app whose live voice path runs its evaluations through Jev
• decider ★ 1.1k - A family of System One-style models fine-tuned from Qwen3.5, designed for one-pass typed decisions with calibrated probabilities
• TypeAR ★ 937 - Type-safe language-model generation and structured decisions
• SREGym ★ 332 integration - When enabled for Codex runs, Jev reviews diagnostic tests and incident-resolution submissions from the supplied evidence
• jev-align ★ 306 - Build calibrated AI Functions from human feedback using Jev and GEPA
• jevbench ★ 240 - JevBench v1 - a benchmark for Jev-class typed decision models: smart, cheap, fast, reliable, open
• tasksource ★ 210 integration - Tasksource recasts hundreds of classification datasets into Jev-compatible typed questions, answers, and procedural training examples
• reflex ★ 163 - A small open decision model: state + typed questions -> calibrated probabilities. A Jev / System One re-creation on Qwen3.5
• Convex Decision Evals ★ 130 integration - Benchmark comparing Jev Choice answers with LLMs on four-option questions about Convex
• open-jev (daseinlabs) ★ 128 - Open Jev implementation with custom finetuning
• Verdict-open-jev ★ 111 - Non-autoregressive decision engine on ModernBERT (151M) with calibrated uncertainty (RLCD), TypeSafe AI Jev benchmark audit, and in-browser WebGPU playground
• jev-curate ★ 107 - High-throughput synthetic and pretraining dataset sifter for TypeSafe Jev. Rust streaming core, Parquet and JSONL I/O, typed Choice/Score/Noul judgments, speculative fan-out, 24.0 rows/sec measured
• WindTunnel ★ 87 integration - WebMCP benchmark comparing browser-agent interfaces, with Jev included as one of the evaluated configurations
• RSI-Jev ★ 86 - Typed-decision models (noul / choice / score) trained by a self-improving loop of AI agents — checkpoints, the code that produced them, and every version that failed
• jevmlx ★ 69 - Jev-style parallel constrained decisions for any MLX model on Apple Silicon. Typed, schema-valid JSON in one forward pass
• open-alternative-jev ★ 67 - Open-source alternative to TypeSafe's Jev: a System One style model layer that gives typed, calibrated decisions from any open-weights LLM in one forward pass (HF + vLLM), with honest benchmarks
• jev-dataops ★ 62 - An open-source JEV-powered workbench for streaming data selection, quality evaluation, automatic LoRA training and held-out model evaluation

Demos & Playgrounds

Small, runnable examples for trying Jev.

• killmyidea ★ 260 - Describe your startup idea. Jev decides: kill it, fix it or ship it

Contributing

See CONTRIBUTING.md for the entry format and requirements.

Know a project that meets the bar? Open an issue or send a PR with the repository link, a category, and one sentence on the decision Jev makes in it.

• Add a project — must satisfy all four criteria above
• Report a dead one — removal is as valid a contribution as addition
• Fix a description — keep it to one sentence

One project per pull request. Bulk submissions of same-day scaffolds are closed without review.

License

MIT License - see LICENSE