TypeSafe Jev skills, agents and MCP servers: open-source projects built on TypeSafe's Jev decision model, security-graded: the jev-ultrafast browser agent, openjev, fast-jev-compaction, Jev code review tools and MCP servers. Vet before you install.
Of the 341 TypeSafe Jev Skills, Agents & MCP Servers here, our pick after running each one's quick start is dzhng/jevgrep (★ 3.2k). To search code by what it does: One npm install, one command to save the key, a doctor check that confirmed the connection, and a search that returned a result. No other key or account needed. See the first runs →
On 2026-10-11 we took 12 of these projects that call the hosted Jev API and gave each the same test in a throwaway sandbox, driven by Claude Code (Claude Opus 5.5): install it the way its README says, then run the README's first example once, with a Jev API key in the environment and nothing else (no GPU, no desktop, no other paid keys, no accounts). The projects do different jobs, so this does not compare how good they are. It answers one thing: can a new user get it running.
A Jev key alone is not enough for these: jev-ultrafast (the most-starred project here also needs a paid text-model key); foreman (needs the Codex CLI signed in); jev-search (needs a Cloudflare account).
Whichever you pick: every project that ran did so on a few Jev calls costing a fraction of a cent, so trying one is cheap. Read what else a project needs before you start.
Listed by GitHub stars. These projects do different jobs, so this is not a ranking of quality: it is whether a new user with only a Jev key gets through the README's first example. One run each, a handful of Jev calls.
What we found: 8 of the 11 we could judge ran their quick start on a Jev key alone, each in about a minute and on a handful of Jev calls. The 3 that did not each need something more: a second paid key, a signed-in Codex CLI, or a Cloudflare account. The most-starred project on the page is one of them.
| # | Project | Quick start | What it needs | Steps beyond the README | Jev calls | Install and run | Evidence |
|---|---|---|---|---|---|---|---|
| 1 | jev-ultrafast ★ 22,555 | Did not run | Needs a second paid key: Jev picked the first action, then the agent stopped for want of a text-model key (TEXT_MODEL_API_KEY). | 3 | 1 | 86 s | The run |
| 2 | fast-jev-compaction ★ 7,570 | Ran | A Jev key only | 4 | 1 | 52 s | The run |
| 3 | jevgrep ★ 2,544 | Ran | A Jev key only | 5 | 11 | 60 s | The run |
| 4 | hermes-jev-skills ★ 1,074 | Ran | A Jev key only | 3 | 1 | 39 s | The run |
| 5 | foreman ★ 732 | Did not run | Jev's routing and assessments came back, but its default workers need the Codex CLI installed and signed in. | 3 | 4 | 71 s | The run |
| 6 | jev-review ★ 680 | Ran | A Jev key only | 3 | 12 | 56 s | The run |
| 7 | jev-seo ★ 547 | Ran | A Jev key only | 2 | 2 | 53 s | The run |
| 8 | jev-search ★ 529 | Did not run | Needs a Cloudflare account: its dev server opens a remote Workers AI session and would not start without one. | 3 | 0 | 103 s | The run |
| 9 | docjev ★ 523 | Ran | A Jev key only | 2 | 1 | 55 s | The run |
| 10 | jev-mcp ★ 516 | Ran | A Jev key only | 3 | 1 | 44 s | The run |
| 11 | JevRouter ★ 457 | Ran | A Jev key only | 3 | 1 | 46 s | The run |
TypeSafe Jev Skills, Agents & MCP Servers are AI-powered software designed to help developers and teams tackle typesafe jev-related tasks more efficiently. These tools are typically published as open-source projects on GitHub and can be integrated into existing workflows via MCP (Model Context Protocol), Claude Skills, or standalone agent frameworks. On Agent Skills Hub, we index 341 quality-scored TypeSafe Jev Skills, Agents & MCP Servers across languages including Python, TypeScript, Rust.
In 2026, the AI agent ecosystem is maturing rapidly. TypeSafe Jev Skills, Agents & MCP Servers can significantly boost development efficiency by automating repetitive tasks, reducing human error, and providing intelligent suggestions. The top 3 tools — jev-ultrafast, fast-jev-compaction, typesafe-computer-use — have earned an average of 344 GitHub stars, reflecting strong community validation. 285 of the listed tools come with clear open-source licenses, ensuring freedom to use and modify.
When choosing among TypeSafe Jev Skills, Agents & MCP Servers, consider these factors: 1) Community activity — GitHub stars and recent commit frequency indicate reliability; 2) Integration method — check if it supports MCP, Claude, or your preferred agent framework; 3) Language compatibility — the most common language in this list is Python; 4) Quality score — Agent Skills Hub's composite score evaluates code quality, documentation completeness, and maintenance activity. Our recommendation: start with jev-ultrafast — it ranks highest in both star count and quality score.
Fastest and cheapest web agent
git clone https://github.com/browser-use/jev-ultrafast.git
Claude Code plugin that replaces the compaction summary with Jev decisions: every tool call and result is scored in one fast request, stale ones are dropped or truncated, everything kept stays verbatim.
npm install fast-jev-compaction
Computer use for about $0.0002 a step: OCR the screen, classify the next action with TypeSafe, click. macOS.
git clone https://github.com/awlevin/typesafe-computer-use
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 multi-tenant trading SaaS with built-in user management, billing, payments, and settlement.
git clone https://github.com/OpenByteInc/QuantDinger.git
Jev-like family of decision models built on top of Qwen3.5/3.8 you can train and run on your own
git clone https://github.com/jaredpalmer/kev.git && cd kev
Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context.
npm install -g @dzhng/jevgrep
A nano replica of Jev: parallel decisions, dynamic candidates, and an end-to-end training pipeline.
git clone https://github.com/TianyuCodings/NanoJev.git
Turn any LLM into a Jev-style decision model: typed decisions, real probabilities, no training. (continue updating, welcome any issue and PR request)
pip install "anyjev[hf]" # library + transformers backend
Jev-compatible API server for open decision models. Ollama for decision models: pull and serve 29 model families (Laya, pplx-decider, Clef, Winnow, coba...) locally or self-hosted behind /v1/systemone. The TypeSafe SDK works unchanged: set its base URL and pick an open model. CLI, desktop app, Docker, MCP server, public bench.
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.
/plugin install reticle@reticlehq
Jev-powered model routing, memory, compaction, skill selection, computer and browser use for Hermes agents (also Claude Code and Codex)
git clone https://github.com/kerpopule/hermes-jev-skills ~/hermes-jev-skills
Ask your Postgres tables questions in plain language. A PostgreSQL extension powered by TypeSafe's Jev.
git clone https://github.com/realZachi/pg-jev.git && cd pg-jev
5–10x faster browser operations: Jev clicks, Codex thinks and verifies. Built at EZCollegeApp.
npx skills add wy-coliney/jev-browser-use -g -a codex -y
MOMO CODE — RSI *JEV*Pi --AI coding agent that evolves with you
git clone https://github.com/momozi1996/momo-code.git
An input that becomes what you mean: one text box that morphs into the right UI as you type. Powered by TypeSafe Jev, works offline.
Run Laya, the open-source Jev-compatible System-1 decision model, from Node.js / TypeScript via ONNX Runtime
npm install @receptron/laya
The open, local take on Jev: typed decisions from an LLM, without generating a single token
git clone https://github.com/Rizzo-AI-Academy/rizzo-flow
The open-source System One decision model. Sub-15ms, non-autoregressive, local drop-in alternative to TypeSafe Jev.
pip install von
JevChat-Windows: Reply helper for chat windows. Offline local OCR reads screenshots; Jev infers intent and offers 3 one-click replies. You send manually.
git clone https://github.com/jev-chat/jev-chat-windows.git
An agent supervisor and software factory foreman powered by decision models
Open, Jev-compatible System One decision server on DiffusionGemma
git clone https://github.com/razorback16/openjev && cd openjev
Live SEO audit for any website from one homepage URL, judged by Jev. PDF, XLSX and Markdown reports.
git clone https://github.com/AgriciDaniel/jev-seo.git
A staged code-review workflow and local dashboard built with TypeSafe Jev.
Route to the cheapest model in claude code for your task using jev-router
npm install -g jev-router
Turn any open model into a classifier/jev endpoint
git clone https://github.com/featherless-ai/simple-jev.git
Astra planner and JEV controller for Minecraft, with native recording, tested routes, and run verification.
Jev: Jupyter Notebook experiments, 3 problem primitives, 18 recipes, voice smart home, model evaluation, fine-tuning, and System One judgment models.
Serve the experimental DiffusionGemma-Jev diffusion model on Cloud Run GPUs (vLLM, RTX PRO 6000), with demo apps and published benchmarks.
Goutou Junshi Chat: a floating window for reading Mac WeChat, analyzing relationships, and drafting replies
Fast, cheap, typed judgments from TypeSafe's Jev model, as MCP tools.
npx -y @jkudish/jev-mcp
Search the web with TypeSafe's Jev: source selection, query understanding and relevance ranking. Built with Search1API.
git clone https://github.com/superagents-lab/jev-search.git
LLM-authored task-specific Jev harnesses with optional full-trajectory reward reflection and GEPA evolution.
/plugin marketplace add https://github.com/TianyuCodings/JevHarness.git
A very fast document classifier/splitter using Jev
git clone https://github.com/jerryjliu/docjev.git
Agent tooling for generative LEGO models building, built with Astra and Opus 5.5, powered by Jev
Tax document page classifier built on Jev decisions. 100% strict accuracy across 261 IRS forms, ~$0.001 per page.
pnpm add github:<org>/tax-doc-classifier
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.
git clone https://github.com/Yinsongxu/LLM2Jev.git
A lightweight Jev-powered router for models, tools, and subagents
npx --yes github:BillionsBobby/JevRouter agent setup --agent all
A TypeSafe/Jev agent that plays Super Mario Bros. from structured emulator state.
py -3.13 -m venv .venv
.venv\Scripts\python -m pip install -e ".[mario,dev]"
$env:TYPESAFE_API_KEY = "your-key"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.
git clone https://github.com/notque/vexjoy-agent.git ~/vexjoy-agent
Control a real browser by voice. Jev (TypeSafe System One) decides intent + target in ~300 ms per spoken word; Playwright acts — often before you finish the sentence.
git clone https://github.com/moritzkremb/jev-voice-browser.git
Open Jev-style typed-decision model that also takes images: photo + app state + typed questions in, calibrated probabilities out, locally.
git clone https://github.com/mohit67890/imajev && cd imajev
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. Zero output-token decoding.
git clone https://github.com/malevrigns/agent-jev.git
WXT browser extension: Jev-powered page clutter removal with reusable template rules.
git clone https://github.com/kitze/unclutter.git
System One MCP connector to evaluate anything fast and cheap. Give your AI agent direct access to models like: Jev, D1, CLM and Laya
Local-first macOS coding workspace with local Laya and optional Jev decision selection
Three AI trading bees on OKX, every decision by Jev. Paper trading by default. Not financial advice.
Crush monitor: use Jev to analyze emotions, intent, and reply performance in WeChat chats. Run locally with your own API key.
EmbodiedJev: MuJoCo robot decision workbench with MiniCPM5-2B, Jev and compatible model APIs
git clone https://github.com/FBddcz/embodied-jev.git
An easy way to use jev with your coding agent for tool calling reasoning
npm install -g jev-gateway
Build calibrated AI Functions from human feedback using Jev and GEPA.
uv tool install jev-align
Adaptive reasoning effort for GPT-6 during Codex tasks, powered by Jev to reduce token usage.
git clone https://github.com/miuuyy/Astra-Ares.git
A self-hosted drop-in replacement for TypeSafe's jev, powered by GliFormer.
Camera-only autonomous drone in MuJoCo with a small judgment model (TypeSafe Jev) in the loop at 2.5Hz
Per-turn model & reasoning routing for Codex, driven by Jev (TypeSafe System One): picks the model, thinking depth and speed mode for every turn.
Jev DSH 决策引擎｜面向 Agent Harness 的结构化决策插件。原生支持 DeepSeek Harness，通过 iPolloWork 支持 OpenCode、Codex Harness。
git clone https://github.com/Devin-AXIS/jev-dsh-decision.git
Ten levels of Jev, from one smart if statement to a coding agent that reaches for Jev on its own
Established strong JEV based natural language to SQL performance and well-covered semantic operators for PostgreSQL with optimized cost and latency. Reusable evidence, with native execution, parallel plans and explainable embeddings in development preview.
Local-first MCP plugin for continuous software-quality review by AI coding agents, powered by Jev.
npx plugins add NiazMorshed2007/jev-review
Reusable project context and specialist workflows for the AI tools you already use. Agents, skills, rules and MCP tools for Claude Code, Codex, Copilot, Cursor and more, plus SDLC workflows, session memory, and Jev decision support that offloads bounded LLM decisions to save tokens.
npm i -g aiwg # install globally
An on-demand crypto market intelligence and decision-support terminal powered by TypeSafe AI's System One model
🚀🚀 A multimodal System One decision model that gives calibrated answers to typed questions about screens, photos, video and text in one forward pass.
pip install git+https://github.com/OmniJev/OneJev.git
The System One harness for System One models. Run Jev and other System One models locally or directly on HarnessRouter.ai.
git clone https://github.com/HarnessRouter/SystemOneHarness.git
Native DeepSeek Harness (DSH) plugin integrating TypeSafe Jev or Decision api like luna as a System One decision layer for agent selection, supervision, corrections, and approvals.
Open-source, local-first social media research agent for Instagram, TikTok, and LinkedIn. Jev routes read-only steps; socai CLI captures cited browser evidence.
Small decision models (jev style) for latency critical scenarios
pip install "jiwo @ git+https://github.com/jiwidi/jiwo"
OpenSource Jev
cd OpenJev # folder name; package is "OpenJev"
npm install
npm run devTypesafe.ai System One Model Jev navigating a Neo4j graph by using a classifier over neighbouring relationships
System 1 decision models (Jev, Laya, Cua-S1) as brain for agents: Browser use, computer use, games and robotics
npx skills add ThinkFlowLab/system1-agents # the skill: Claude Code, Codex, Cursor
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.
git clone https://github.com/agent-labs-dev/fastbrowse.git && cd fastbrowse
A small open decision model: state + typed questions -> calibrated probabilities. A Jev / System One re-creation on Qwen3.5.
git clone https://github.com/kshetrajna12/reflex
Research Instagram Reel scripts with Apify, Jev, and Fireworks or Groq. Bring your own keys, compare hooks, and record the patterns.
git clone https://github.com/artemnovitckii/creator-lab.git
TypeSafe Jev as a decision layer for the Pi coding agent: a measured tool-call gate plus jev_ask for typed, calibrated answers
pi install npm:@y0usaf/pi-jev安卓AI聊天端:开箱即聊，无损导入酒馆卡，缓存强省用量，Jev 决策，专属RP模式，防空回，主动信息，前端卡支持，高定制主题，QQbot，AI群聊，插件系统，强兼容中转站。手机移动端原生支持，求 Star 收藏⭐Android AI Chatbox: lossless SillyTavern card import, aggressive cache optimization, Jev decisions, multi-dimensional memory, anti-blank-reply, proactive messages, AI group chat, SillyTavern themes, plugin system, Native mobile.RikkaHub
Claude Code plugin: trim long Bash output with TypeSafe Jev before the model sees it
git clone https://github.com/tamaratran/jev-pruner.git
A skill for writing and improving programs that call Jev, TypeSafe's System One model
/plugin marketplace add dbreunig/building-with-jev-skill
AI-assisted Windows and Chrome GUI delegation for Codex: task contracts, local execution, Jev semantic decisions, recovery and outcome verification.
git clone https://github.com/YUTA-fywoo/jev-gui-delegate.git C:\jev\jev-bridge
An open training and inference stack for Jev-style decision models. Train models to score dynamic candidate branches from a shared prefix, with support for high-cardinality choice, calibration, and fast batched inference.
Smart web search CLI for agents, backed by Jev. Saves a lot of tokens.
brew install dorkitude/webctl/webctl
JevK5: open-weight alternative to TypeSafe Jev. Typed decisions with probabilities in one forward pass; Apache-2.0 weights and code.
pip install "jevk5[fast] @ git+https://github.com/allebee/jevk5@v0.2.0"
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 (TTS/STT/realtime), RAG, memory, file processors. Powers Tara, Yama and Clairvoyance at Juspay.
npm install @juspay/neurolink
A proactive pi-durable news agent that sweeps on a cron tick and ranks every story as Hype, Facts and Signal with jev. Runs on your Mac, an always-on Mac mini or a VM.
grep by meaning, across languages. TypeSafe Jev scores every line against a meaning; combine meanings with AND/OR/NOT. 意味で探す grep。日本語で英語を、英語で日本語を検索できる
npm install -g @uehaj/sys1grep
Review behavior, not just diffs. Jev prioritizes human attention; OpenAI explains the changes. Local CLI + agent skill + GitHub extension.
git clone https://github.com/egma-ai/jev-code-reviewer.git
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, Cursor, pi, OpenCode, ACP.
npm i -g jev-guard
Read Aloud — R2T2 streaming ASR listens, Jev judges word by word, and scores English reading.
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.
npx github:UditAkhourii/quicksilver
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.
uv tool install jev-grep # the command it installs is jgrep
Unofficial TypeSafe Jev showcase — System One decisions, not chat.
git clone https://github.com/cobusgreyling/Jev.git
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.
npm install -g jevmem
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.
git clone https://github.com/Dicklesworthstone/skillranker.git
an open-source, from-first-principles reconstruction of the ideas behind TypeSafe AI's Jev, written in pytorch
git clone https://github.com/kyegomez/open-jev
Talk to your Mac. Local whisper.cpp + one Jev (TypeSafe) call per command + macOS automation.
Connect TypeSafe Jev to Grok Bot as a cheap decision layer - usage gates, skill template, examples
git clone https://github.com/Bodila51/grok-bot-jev.git
Jev, an AI decision model, plays Pokémon Red. It beat the game in 37h 40m.
git clone https://github.com/christianmat/jev-pokemon && cd jev-pokemon
Browser automation where an LLM plans and Jev (Typesafe System One) decides. Library, CLI and MCP server.
git clone https://github.com/Ying-Kai-Liao/jev-browser && cd jev-browser
Say it, and your Mac does it. A computer-use harness on Jev that reads the screen through Accessibility. Fast, no vision model
bash build.sh
open "$HOME/Applications/Desktop Voice.app"A linter to codify code taste using Jev
go install github.com/codegirl-007/jevlint/cmd/jevlint@latest
lint text in code by jev scorerer
npx -y jev-lint check src --dry-run # what it would ask, and the price. No request.
A lightweight, native Swift 6 bridge integrating TypeSafe AI's Jev System One decision model into Apple's Foundation Models framework.
CLI that picks Cursor, Claude Code, Codex, or OpenCode + model/effort for a task, then launches it. Powered by Jev and Herdr
git clone https://github.com/nidhi-singh02/agent-router.git
Jev: Choice/Score/Noul convert text to typed judgments/probabilities for classification/scoring/routing; model comparison, CSV/Excel, replay, offline
git clone https://github.com/NanmiCoder/jev-arena.git
Semantic grep: find code by describing what you're looking for, powered by Jev.
cargo install jegrep
Fish-style zsh history autosuggestions ranked by Jev (TypeSafe)
git clone <this repo> ~/.zsh/jev-shell-history
cd ~/.zsh/jev-shell-history && npm installBounded TypeSafe Jev workflows for coding agents.
git clone https://github.com/devagrawal09/stanley-code.git
Your browser. Ready to act. An MIT-licensed desktop browser agent for website tasks, cited research, and supported games.
git clone https://github.com/michaelswissa/jevry.git
Detect youtube sponsor segment with live audio and transcript powered by Jev
Jev-style structured decisions on DiffusionGemma: the example server from vLLM PR 57250
Non-autoregressive decision engine on ModernBERT (151M) with calibrated uncertainty (RLCD), TypeSafe AI Jev benchmark audit, and in-browser WebGPU playground
git clone https://github.com/Heman10x-NGU/Verdict-open-jev.git
Free open source alternative for Jev that supports text, documents and images and runs at ~2GB of ram on a CPU.
git clone https://github.com/deepanwadhwa/OpenDecision.git
Laya vs Jev: local MLX and hosted AI decisions playing T-Rex side by side, with live metrics and replay recording
git clone https://github.com/virajbhartiya/laya-vs-jev.git
The model router for AI agents.
git clone https://github.com/thinkany-ai/autojev.git
Put a live Jev (TypeSafe) meter on any video: every sentence scored, rendered as a 16:9 edit
git clone https://github.com/ChetasLua/jevmeter.git && cd jevmeter && ./install.sh
A self-hosted API and web interface for Laya’s structured decision models, compatible with the TypeSafe Jev API format.
docker run -d --name laya-server --init --restart unless-stopped \
-p 8080:8080 \
-v laya-data:/data \
-e LAYA_ADMIN_USERNAME=admin \
-e LAYA_ADMIN_PASSWORD='change-this-admin-password' \
1panel/laya-server:latestJev-powered semantic code search for coding agents — find behavior across repositories via CLI or MCP, with exact source excerpts and line numbers.
npm install -g @nassim-arifette/jevgrep
Open-source AI email triage for Gmail. Sorts your inbox into Needs reply, Updates, Promos, Sales and Spam with Jev, TypeSafe AI's decision model, via Vercel AI Gateway. Read-only, runs locally, 1,000 emails in about a minute for 3 cents.
git clone https://github.com/fazlerocks/jevmail
TypeSafe Jev for DeepSeek Harness, the Model Context Protocol, and plain Node: typed judgments instead of prose, offline by default.
npx -y jevcore-mcp
AI-driven intraday trading bot for Indian stocks. Jev ranks the Nifty 50 every 15s; code sizes each trade and places the stop; orders go live through Zerodha Kite or Kotak Neo. Day replay, kill switch, daily loss halt, terminal dashboard.
Turns Jev into a chatbot
poetry installA small, type-safe client for asking AI questions about your data, powered by TypeSafe Jev.
npx skills add pithings/advocaat
🧹 Fun project: a Chrome extension that asks a tiny AI decision model (TypeSafe Jev) "is this DOM element an ad?" and pops it off the page. BYOK, no backend, not a real ad blocker.
Everything you need to run TypeSafe's Jev with Claude Code: a tool-call guard, tier guard, file search, browser agent, review, belay, compaction and installers.
git clone https://github.com/jonathanavis96/jev-kit.git ~/code/jev-kit
Faster inference backend for Open-Jev-27B: fused CUDA kernels, prefix tree, CUDA Graphs (B300, bf16)
Fine-grained robot control with Jev, physics previews, and configurable LIBERO tasks.
git clone https://github.com/Dimweaker/jev-libero.git
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.
Typed semantic decisions for Unix pipelines and CI, powered by TypeSafe AI Jev.
pipx install https://github.com/sharziki/semdecide/releases/download/v0.2.1/semdecide-0.2.1-py3-none-any.whl
Pre-alpha PostgreSQL extension for TypeSafe AI (Jev) categorical classification
git clone https://github.com/giuliosmall/pg_typesafe.git
MCP server and CLI that crawls websites for AI agents: finds the answer page 2x more often than Crawl4AI, 10x cheaper per answer than letting Claude crawl.
A decision cache for TypeSafe Jev-class models — memoize decisions so repeats are free, deterministic, and shareable. One 2 MB binary.
Jev-style parallel constrained decisions for any MLX model on Apple Silicon. Typed, schema-valid JSON in one forward pass.
pip install git+https://github.com/bnsd55/jevmlx
JEV-inspired parallel decisions for CUDA LLMs. One context, many decisions. vLLM API, game-agent examples, and reproducible benchmarks.
git clone https://github.com/kikoncuo/jevfire.git
Meet JevEmbed — turn embeddings into decisions. Choose, score, and judge with your choice of embedding model.
Local proxy that learns your app's typed LLM decisions and answers them with a Laya head. Jev and OpenAI compatible.
pip install "stuntd[train]"
Semantic tool routing and typed System One decisions for the Pi coding agent using TypeSafe Jev
npx pi-jev-gate -c "All exported functions have TypeScript type annotations" --diff
Distill a repeated Jev classification task into a local model, on the fly — same answers, your hardware.
docker run -d -p 8080:8080 -v jevstiller-data:/data ghcr.io/tomerglick57/jevstiller
WeChat (FkWeChat plugin): Long-press a message to analyze intent, emotion, and reply stance. Only you see the prompt.
OpenJev: open-source, local alternative to Jev. System-One decision models: typed questions in, calibrated probabilities out; general decisions and browser-agent steps.
Best Roblox Studio Instance Reviewer 2026: Typed Jev Criteria and Exact Citations
Claude Code / Codex / pi plugin that hands agent steps needing no text output to Jev (TypeSafe's judgment model) — measured p50 ~230 ms and ~$0.02 per 1,000 judgments, with typed escalation back to the LLM
npx -y jev-use install # wires Claude Code, Codex, and pi — whichever it finds
An open-source JEV-powered workbench for streaming data selection, quality evaluation, automatic LoRA training and held-out model evaluation.
git clone https://github.com/RenaGao/jev-dataops.git
An unofficial local visualizer for Jev (TypeSafe): a live view of every call your code makes. Not affiliated with TypeSafe AI.
MCP server that puts TypeSafe Jev on the coding loop in Cursor, Codex, and any MCP client
Jev picks which of your rules apply to each prompt, so Claude only sees the ones that matter.
/plugin marketplace add EliaAlberti/jev-rules
git clone https://github.com/openroboto-ai/jev-robot-control.git
Two-stage JEV control of a Franka Panda in MuJoCo
git clone https://github.com/lykycy123/RoboJEV.git
🚀🚀 A 0.8B JEV-like multimodal model playing GUI games directly from raw pixels.
git clone https://github.com/OmniJev/PlayJev && cd PlayJev
grep for what code does, not what it's called. Semantic code search powered by TypeSafe Jev.
npm i -g jevgrep # installs the `jgrep` command
Generation-free typed decisions on Indic LLMs. An open Jev-style inference engine on sarvam-1: constrained logit readout instead of autoregressive JSON. Runs client-side in the browser.
Can a model design a dashboard? A console whose whole design system is generated at runtime by Jev from a one-sentence brief.
A local, API-compatible replacement for Jev, TypeSafe's System One model
Ruby client for decision models such as Typesafe Jev
gem "ruby_decision_model"Jev, TypeSafe's System One classifier, as a tool inside Claude Code, Codex, Pi, and OpenCode: typed classify, check, score, rank, and ask, plus one-command setup.
npx -y @francoischastel/jev-code setup
Supersimple way to serve LLMs as a Jev-like endpoint
uv tool install jevify # or: pipx install jevify
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* traversal for any graph database.
git clone https://github.com/bodepudimuneendra-netizen/ConsciousDB.git
A Jev powered LinkedIn recruiting agent. Watch it browse relevant profiles, save links, and review evidence against your hiring brief.
git clone https://github.com/skeptrunedev/jev-recruiter.git
Route HTTP requests by meaning. A semantic router for Hono powered by Jev.
npm i hono-jev-router
Sniff out the right record: Jev-powered search for Rails with index-time labels, query understanding, and streamed reranking on top of your own keyword and embedding search
Jev-powered relevance filtering and reranking for RAG in Python.
uv add jev-reranker
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.
Go SDK and CLI for TypeSafe Jev: typed decisions (yes/no, choice, score) from a model
Claude Code Stop hook that blocks an unverified done: reads the transcript for evidence, asks Jev once, fails open on everything else
open-jev is a browser-focused TypeScript library for typed decisions: one piece of text (the state) plus any number of typed questions go in, and one forward pass returns a calibrated probability distribution per question. Nothing is generated, so an answer is always one of the options you provided.
Self-evolving cognitive decision engine & TypeSafe Jev alternative. Deliberative decision flywheel ('explore first, crystallize later' System 2→1) with 100% option-order invariance. Typed probabilistic API (Choice, Noul, Score) for Open LLMs, Laya (ModernBERT), & commercial Jev with calibrated logprobs and adaptive safety guards.
Jev-native probability tree and graph runtime for verifiable multi-step decision making.
Command-line tool for TypeSafe AI's Jev model. Ask yes/no, multiple-choice and rubric questions about any text and get calibrated probabilities back. Answers become exit codes for shells and CI, JSON for scripts, and MCP tools for AI agents.
Text extraction with Jev: field extraction, PII detection and exact quotes, built on TypeSafe's Jev.
Practical, tested recipes for TypeSafe's Jev decision model on OpenRouter: support triage, database indexing, file organizing, tagging, taxonomies, dedupe, PII detection, extraction, search re-ranking and a browser agent.
Score every test against a git diff with Jev, and emit the filter arguments vitest, node:test, Playwright, cargo test and go test already understand
离线可用的本地类型化决策：4 核 CPU 单题 15.6ms。Local & offline Jev / System One inference on CPU — ONNX + INT8, no torch at runtime. 支持 laya / kev / PlayJev
Serve typed-decision (System 1) models — Laya or your own — on NVIDIA GPUs or Apple Silicon, with a Jev-compatible API and coding-agent integrations
用 TypeSafe Jev 推荐已安装 Skill / Bounded installed-skill recommendations with TypeSafe Jev. Python CLI, Codex skill, bilingual docs and live examples.
Claude Code compaction in ~0.5 s instead of ~35 s: Jev keeps what's still needed word for word and moves the rest to a drawer file. Fork of fast-jev-compaction.
Calibrate Jev questions against your own labels: tune criteria on labelled examples, confirm on a held-out set, get a verdict per question. Unofficial.
Skim pages, files, and lists with Jev: keep only what helps with your goal. Saves agent tokens. Built with go/cobra.
A chatbot built on a model that cannot generate text (TypeSafe AI's Jev, driven autoregressively)
Voice control for macOS that acts partway through your sentence, built on Jev (TypeSafe)
Pluggable decision layer for the ego lite browser: one System One (Jev) call per step replaces the per-step LLM turn, and the backend can be swapped for a local OpenAI-compatible model. Fail-closed execution guards. The measured one — raw bench data, 16 suites, changelog with corrections.
sort by meaning: order lines along a plain-English dimension, from pairwise comparisons judged by TypeSafe's Jev model
A tiny jev-like model that answers Choice, Score and Noul questions in one forward pass and returns calibrated probabilities. MLX or PyTorch, fully offline, System One compatible.
Just Jev it. A Jev-first agent loop with reusable actions for DeepSeek Harness.
Jev-like decision models that can also see images — open weights on Qwen3.5, run on your own GPU
Unofficial study: Jev-style parallel typed decisions on stock 1.5B-8B models on an Apple Silicon laptop. Benchmarks, research notes, and a Hugging Face Space demo.
MCP server for TypeSafe Jev: typed classify, score, check, match and screen for any agent, with confidence on every answer
Super fast Jev like server in JS, model agnostic, working with any OpenAI compatible endpoint
Run Jev-style typed decisions locally on your Mac with low RAM usage and fast responses
A local, offline System One server compatible with TypeSafe's Jev API. It answers typed yes/no, category and score questions with small open models.
The invalidation layer for AI memory. Every fact gets a lease; new evidence ends it. Built on TypeSafe Jev.
BrighTO LLM Router: free open-source, ultra-fast Rust LLM gateway/router and LiteLLM/Bifrost alternative with OpenAI + Anthropic APIs, Claude CLI bridge for OpenAI-compatible models, Qwen3.8 dual protocol, model load balancing, fallback routing, team API keys, token budgets, SystemOne decisions for Ollaya/Laya, JEV/DJEV and Quyết.
System One judgments (noul/choice/score) from any LLM in one prefill — a Rust, Jev-compatible /v1/systemone engine
Command Code mod: nudges the agent to keep going when it stops with work left, judged by Jev
Hand a whole browser task off in one call: a decision model drives the page server-side, so a flow costs one call, not a turn per click. Ref-based element tables, code-checked assertions, zero-model macro replay, over the Chrome DevTools Protocol.
Use Jev (TypeSafe's System One model) as a calibrated reranker: one call, up to 30 documents, a probability per document. Apache-2.0.
Jev-shaped (TypeSafe System One) classification wrapper over OpenAI-like clients
Train JEV-style System 1 decision models on Qwen3.5 and Qwen3.8-27B: 24-slot decision head, KL distillation, calibration and serving. 在 Qwen 上训练 JEV 风格的 System 1 决策模型。
JEV-powered market decision bot for stocks, crypto and memes. State in, BUY/SELL/HOLD/AVOID out, paper by default
pre-commit hook: one Jev call judges whether your commit message matches the diff, plus debug leftovers, scope creep, and a secret belt
Simulate how Indonesian netizens react to a post before it goes live (TypeSafe/Jev)
⚡ Adaptive multi-model LLM router — 80+ providers, Jev System One single-pass routing (model=jev-auto), pheromone-trail failover, parallel ensemble merge. npm: adaptive-memory-multi-model-router
Semantic "Did you mean?" for any CLI — wraps commands and uses TypeSafe's Jev to match subcommand typos by intent, not edit distance.
Typed-judgment admission control at the traffic edge: three-layer prompt-injection and abuse filter for nginx/OpenResty, powered by TypeSafe Jev. Fail-open, cached, hot-reloadable.
Jev for every agentic development workflow — the System One decision model wired into whatever harness an agent codes in: OpenCode today, Claude Code and Hermes adapters next.
A stronger one-pass scorer over a variable list of text options. Hashed n-gram encoder, rival-aware attention, gated head, temperature scaling — with a head-to-head benchmark vs the jevlike starter design.
Decision harness for TypeSafe Jev — confidence gates, shadow mode, recipes, and evals. Claude CLI 48.9s → Jev 1.3s on the same row-filter job.
Pi coding-agent extension: TypeSafe Jev checks for tool calls, tool outputs and replies (prompt injection, approvals, secret scrubbing, task pinning)
Android accessibility WeChat interpreter: screen-reading only, 3 Toasts (intent/emotion/urgency/suggestion), no reply generation or sending, no third-party dep…
An unofficial, provider-neutral Unix client for Jev from TypeSafe AI. Written in Go.
openvons (open-Jev): A decision layer that answers finite choices probabilistically through text, images, and Japanese voice commands.
Docling → Jev vs Docling → Gemini 3.8 Flash vs Gemini reading the PDF: same documents, one clock, scored against arXiv's own metadata
Jev-assisted retrieval and evidence-preserving inspection for local Markdown, Obsidian vaults, and Logseq Markdown graphs.
Eyes are All JEV Needs - real time visual devisions from RGB, video and RGB-D cameras.
Plain-English browser e2e tests cheap enough to run on every PR. Built on Jev and Playwright, open source, bring your own key.
unofficial go sdk for typesafe ai. not affiliated with or endorsed by typesafe ai. a side project built to fill the missing go sdk gap, for the community to use.
Coding-agent tools on TypeSafe's Jev, measured in public. review-router: caught 13 of 13 CVE fixes a filename rule sent to a quick review. jev-gate: a strict second lock, 0 of 12 dangerous test commands ran where Claude Code's auto mode let 8 run.
Per-row semantic judgements over large relational tables on a decision model (Jev)
CLI-first field-level anomaly detection for structured files and text, powered by Jev
Simplified RL for Calibrated Decisions: parallel constrained JSON decoding + GRPO with proper-scoring-rule rewards + calibration eval (Jev/RLCD reconstruction)
MorrowCache: open-source semantic cache for OpenAI-compatible LLM APIs. Local proxy that reuses answers for paraphrased prompts and AI agent retries. npm: @kushalicious/jevcache
Hermes Agent plans, Jev (TypeSafe) picks bounded actions, Mineflayer executes: Minecraft with no screenshots or keypresses from a model. Includes the reproduction of rmalde/minecraft-agent's Ender Dragon run.
📏 Claude Code plugin that makes Claude follow your AGENTS.md: every reply and edit checked by TypeSafe Jev ✅
Jev 多模态手机自动化：面向 Android、iOS、HarmonyOS 的独立 Python 点击 SDK，毫秒级本地处理。 Jev-powered multimodal mobile automation: a standalone Python tap SDK for Android, iOS and HarmonyOS, with millisecond-level local processing.
使用 TypeSafe Jev 审查 Skill 与 MCP 可疑行为 | Review Agent Skills and MCP code with Jev, static evidence, and explicit coverage gaps
Typed decisions from the shell: an unofficial stdlib-Python CLI and Agent Skill for TypeSafe's Jev model, via the TypeSafe API (default) or OpenRouter. Yes/no, choice and ordinal scores with calibrated probabilities, semantic grep and batch mode.
Open-source, local Jev alternative: a System One decision layer with provably calibrated confidence (ECE 0.170→0.071). Typed decisions, runs offline, no API key, no waitlist.
Coding-agent memory where every fact is a verbatim quote graded by TypeSafe Jev's calibrated confidence. Local-first, SQLite receipts, zero dependencies.
⚡ 100% offline, zero-cost drop-in alternative to TypeSafe Jev System One decision models in Rust (<18ms on CPU)
Recursive Jev choice over a taxonomy. Select from more than 255 options without breaking TypeSafe Jev's choice cap.
Stop guessing confidence thresholds: calibrate, threshold, and drift-check typed decision models (TypeSafe Jev) against an LLM teacher.
Local Jev-compatible evaluation server: POST /v1/systemone with typed noul/choice/score, open weights, no waitlist
Jev-powered traffic signal control on CityFlow with structured phase and green-time decisions.
Can we stop AI agents from going rogue before they take over the world? Immune Harness is a security layer for AI agents. It checks every tool call before it runs and blocks the risky ones. When Jev catches a new attack, the harness writes a new policy, or rewrites an existing one to cover the new variant, and tests it before it goes live
Native Swift implementation of OpenJev, the Jev-compatible typed-decision server: yes/no, choice and score answers read from model probabilities, on Apple silicon Macs and in iPhone apps. Serves Verdict, Laya, DiffusionGemma and JevK5. Apache-2.0.
Ask a yes/no question of every function in a codebase. Ranked answers in seconds, for cents. Grep whose pattern is a question, powered by TypeSafe Jev.
Escolhe o modelo certo pra cada mensagem no Hermes, usando o Jev da TypeSafe. Codex e Claude via assinatura.
Read your LinkedIn saved posts into a filterable board, classified by Jev.
Elixir client for System One decision models: TypeSafe Jev, Cloudflare Clef, OpenRouter, self-hosted (CLM). Full probability distributions, behaviour + stub, retries, telemetry.
Connect JEV to MCP clients and compare its judgments against general-purpose LLMs using shared datasets and measurable accuracy.
LlamaIndex reranker + router powered by TypeSafe Jev - typed scores/choices, cheaper than LLM-as-judge.
Code-review gate for CI and coding agents: asks TypeSafe Jev small typed questions about functions, files, tests and docs, and reports findings with locations and probabilities
Compile agent policy prose into deterministic verdict programs: narrow evidence questions for the model, the verdict computed in code. Install: npm i -g jev-compiler
Jev Browser — indexed browser automation. Jev chooses the control, Playwright acts. A CodexQA skill.
A text box that tells you how your message sounds before you send it. Powered by Jev.
Pytest plugin for semantic assertions on LLM output using Jev's calibrated probabilities.
Fast, cheap, probability-scored decisions for coding agents, scripts and CI. CLI + MCP server + Claude Code plugin for the OpenAI Decisions API and Jev (TypeSafe AI).
A word-level language model whose output layer is Jev: n-gram drafter, Noul chunk verification, bits-per-token eval
A small, fast prose linter: ruff-style rule codes for writing, backed by TypeSafe's Jev model
Fast, cheap judgment for AI coding agents: semantic search, focused reads and list picking in ~2s. CLI + MCP server on TypeSafe Jev. Benchmarked on SWE-bench.
Pipe anything into Jev, get typed decisions out. A Unix filter that lets agents offload bulk judgments to a System One model.
Open decision engine: type-safe answers with raw softmax probabilities from a local LLM or any OpenAI-compatible API. Crush radar for WeChat/QQ/Telegram chats - verdict + next-move advice. | 开放决策引擎: 本地 LLM 或任意 OpenAI 兼容 API, 原始 softmax 概率 + 类型安全回答. 微信/QQ/Telegram 追人雷达: 冲/稳/缓/停 + 下一步建议.
Jeff 1 — local open-weight typed decision / fact-checking model (Jev-compatible)
Jev reads a whole novel in seconds. Every passage becomes a row of colour.
Fast Rust CLI for TypeSafe Jev: typed decisions, offline linting before you pay
Make room for useful evidence. Inspectable context selection for RAG, with Jev reranking and open benchmark studies.
An Android assistant using JEV structured assessment and local rules to review salary, benefits, hidden risks, source text, and HR questions on BOSS Zhipin job…
Fast deep research for the pi coding agent: reads up to 100 pages in full per round, Jev keeps only the passages that answer your questions.
A Stop hook that stops your coding agent from stopping too early. Plain-language rules, judged by jev.
Fast typed AI decisions on live crypto markets using TypeSafe AI Jev.
Ask Jev whether a sleeping agent's wakeup is worth a full LLM turn before you resume it. A fail-open wake gate for long-running agents on Workers, Durable Objects and Node.
Self-contained, ultra-fast System 1 decision engine with WebGPU/WASM acceleration, 100% wire-compatible with TypeSafe Jev (/v1/systemone).
Japanese System One decision model (Jev-style): typed answers and probabilities in one forward pass, no text generation. Includes bench_ja/bench_en and a Laya position-bias repro.
Go client for TypeSafe's System One API and its model Jev: typed judgments and calibrated probabilities instead of generated text
Calibrated alignment verifier for LLM responses and agent plans — powered by Jev
Fast semantic code search powered by Jev. Find the relevant files, line ranges, and snippets
Catch breaking API behavior hidden in OpenAPI prose with deterministic checks and TypeSafe JEV System One semantic review.
Audit your git diff against YAML coding-standards packs using TypeSafe's Jev model, from a CLI or your AI agent's command/skill.
Grok Build with TypeSafe Jev routing tool selection once per turn: 22 to 40% cheaper on the same tasks
Let Claude code, chatgpt codex or control your own Chrome. Chrome extension + MCP server: Jev, TypeSafe's decision model, picks each click in ~0.5 s for a fraction of a cent. MIT, bring your own OpenRouter key.
The simplest way to write decision workflows in Python. Python with a smart if.
Dynamic Jev: a Jev-style typed decision API on any stock open-weight LLM. Each question compiles its own readout head at request time; no training, no generation for closed answers.
cuda-Jev — a CUDA-native Jev System One decision inference engine. Jev compatible API, examples, and reproducible benchmarks.
Paste anything, get the right tool. JSON, JWTs, cron, stack traces and more. Powered by Jev.
Use the TypeSafe Jev decision model (System One) anywhere: zero-dependency OpenAI-compatible bridge for 9Router, Claude Code, Cursor, Cline & any OpenAI SDK. Typed yes/no, choice & score judgments via CLI or HTTP.
Exact token counting and request-cost prediction for TypeSafe's Jev (tiktoken-style)
Ask a question about a codebase, get the files that matter. Live, parallel file relevance with TypeSafe's Jev.
ComfyUI nodes that branch workflows on natural-language judgments from System One models (local Laya or hosted Jev)
Learn decision models by building with them. TypeSafe Jev and Microsoft Decision-1 route support tickets in under a second for about $0.02 per 1,000 calls; Claude handles the hard cases. Interactive site, benchmark, calibration tests and docs.
Local JEV-style decisions with DiffusionGemma on Apple Silicon, with benchmarks and coding-agent examples.
Typed decisions (choice/score/noul) from one encoder pass - faster, better-calibrated successor to Laya, Jev wire compatible
Generative UI sandbox: Jev composes the interface as you type, one typed decision at a time
Real-time risk triage gateway for exchange deposits and withdrawals — Jev (TypeSafe System One) handles triage only; adjudication stays in deterministic code.
| Tool | Stars | Score |
|---|---|---|
| jev-ultrafast | ★ 24.2k | 81 |
| fast-jev-compaction | ★ 7.8k | 75 |
| typesafe-computer-use | ★ 1.3k | 66 |
| QuantDinger | ★ 13.5k | 75 |
| kev | ★ 9.8k | 78 |
| jevgrep | ★ 3.2k | 69 |
| NanoJev | ★ 2.6k | 66 |
| AnyJev | ★ 1.6k | 64 |
| ollaya | ★ 1.4k | 67 |
| jevlike | ★ 1.4k | 58 |
| reticle | ★ 1.3k | 67 |
| hermes-jev-skills | ★ 1.3k | 70 |
| pg-jev | ★ 1.3k | 70 |
| jev-browser-use | ★ 1.2k | 66 |
| momo-code | ★ 1.1k | 75 |
| shapeshift | ★ 1.1k | 60 |
| laya | ★ 1.0k | 70 |
| rizzo-flow | ★ 956 | 60 |
| vllm-jev | ★ 943 | 59 |
| von | ★ 922 | 69 |
The most-starred are browser-use/jev-ultrafast (a web agent), OpenByteInc/QuantDinger (a trading system with Jev integration), jaredpalmer/kev (a Jev-like family of decision models you can run yourself) and tamaratran/fast-jev-compaction (a Claude Code plugin that compacts context with Jev decisions). Every repo on this page is also in the open-source list zhuyansen/awesome-typesafe-jev on GitHub.
Through TypeSafe's own API (POST api.typesafe.ai/v1/systemone with model jev-latest) or through OpenRouter. Projects under SDKs, MCP & APIs on this page wrap it for other software, for example realZachi/pg-jev, a PostgreSQL extension that answers plain-language questions about tables.
The top typesafe jev tools in 2026 are jev-ultrafast, fast-jev-compaction, typesafe-computer-use. Agent Skills Hub ranks 341 options by GitHub stars, quality score (6 dimensions including completeness, examples, and agent readiness), and recent activity. The list is rebuilt every 8 hours from live GitHub data.
jev-ultrafast (24.2k stars) is the most adopted choice for general typesafe jev workflows, written in Python. fast-jev-compaction (7.8k stars) is a strong alternative and uses TypeScript instead. Pick by your existing stack: match the language and runtime your team already uses to minimize integration cost. If unsure, start with jev-ultrafast — it has the deepest community and the most examples online.
Avoid pre-built typesafe jev tools when (1) your use case requires deep customization that the tool's plugin system doesn't support, (2) you have strict compliance requirements that ban third-party dependencies, (3) the tool's maintenance is inactive (last commit >6 months ago), or (4) your data volume is small enough that a 50-line custom script is cheaper than learning the tool. For most production workflows above 100 requests/day, the time savings from a maintained tool outweigh the customization loss.
TypeSafe Jev focuses specifically on typesafe jev skills, agents and mcp servers: open-source projects built on typesafe's jev decision model, security-graded: the jev-ultrafast browser agent, openjev, fast-jev-compaction, jev code review tools and mcp servers. vet before you install. OpenAI Codex Skills is a related but distinct category — see https://agentskillshub.top/best/codex-skills/ for those tools. The two often appear in the same agent pipeline but solve different problems: choose typesafe jev when your primary goal is the specific task, and openai codex skills when the workflow is broader.
For most teams, yes. jev-ultrafast has 24.2k stars worth of community testing, handles edge cases you haven't thought of, and ships with documentation. Build your own only when (1) your requirements are deeply non-standard, (2) you have a security/compliance reason to avoid OSS dependencies, or (3) the maintenance burden is small enough (<200 lines of code) that you'll save time long-term. The break-even point is usually around 2-3 weeks of dev time saved.
Most typesafe jev tools listed are open source under permissive licenses (MIT, Apache 2.0). A handful offer paid managed/cloud versions on top of free self-hosted core. Always check the LICENSE file on each tool's GitHub repository before commercial use — some use AGPL or non-commercial restrictions that may not fit your deployment model.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy: 42 published rules, covering credential harvesting, data exfiltration, curl | sh installers and more. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: