Project directory
GitHub projects using Jev AI
A directory of 156 public GitHub projects that call TypeSafe’s Jev AI, modeled on the jev101.com tool index. Blurbs are directory summaries, not reviews.
Source lists: cobanov/awesome-jev and fatwang2/awesome-jev (indexed 2026-09-19). For the bilingual encyclopedia, see jev101.com.
Official SDKs
JStypesafe-sdk-jstypesafe-ai/typesafe-sdk-jsOfficial TypeScript and JavaScript SDK with inferred answer types.PYtypesafe-sdk-pythontypesafe-ai/typesafe-sdk-pythonOfficial synchronous and asynchronous Python SDK.SYsystem-one-adapter-pythontypesafe-ai/system-one-adapter-pythonDrop-in adapter for comparing the System One interface with LLM providers.Sskillstypesafe-ai/skillsOfficial agent skills for building and evaluating System One workflows.
SDKs and clients
ADadvocaatpithings/advocaatSmall type-safe client for asking Jev questions about datasets.JEjevdannote/jevElixir/OTP client designed around GenServer replies and pattern matching.JEjev-axishiftynick/jev-axiCLI for picking, rating, checking, ranking, triaging, and guarding from the shell.JEjev-dslinanna-malick/jev-dslEarly-alpha Haskell DSL that encodes typed question packets and decodes answers; HTTP transport is left to the caller.JEjev-mcpblakestone-x/jev-mcpMCP server exposing classify, score, check, match, and screen tools.JEjev-mcpBYK/jev-mcpAn eval-first MCP server for Jev, that returns typed judgments (noul, choice, score) with probabilities instead of generated text.JEjev-shell-historymrnugget/jev-shell-historyRanks existing zsh history entries for inline completion; accepting a suggestion does not execute it.JEJevbridgegamesonrblx/JevbridgeACP/MCP adapter for using Jev alongside coding and chat models.JEjevclientAboveColin/jevclientAsync Python client for typed Jev questions and probabilities.JEjevrsimxnherrera/jevrNative R client for typed questions and provider-independent answers through TypeSafe or OpenRouter.LAlaravel-typesafe-jevButochnikov/laravel-typesafe-jevLaravel integration with typed responses, async requests, and testing fakes.RUruby_decision_modelobie/ruby_decision_modelRuby client with standard-library transport for TypeSafe and OpenRouter decision endpoints.SEsemdecidesharziki/semdecideTyped semantic decisions for Unix pipelines and CI.TYtypesafe-gozhirschtritt/typesafe-goIdiomatic Go SDK for the TypeSafe API.TYtypesafe-mcpitsmostafa/typesafe-mcpMCP connector that gives agents access to Jev decisions.TYtypesafe-sdk-javaPremo-Cloud/typesafe-sdk-javaCommunity Java 17 client for Choice, Score, and Noul, with an optional Spring Boot starter.ZOzod-jevjomatsu/zod-jevPairs local Zod shape validation with Jev semantic validation.JEjev-goGaurav-Gosain/jev-goGo client for TypeSafe's System One API and its model Jev: typed judgments and calibrated probabilities instead of generated textJEjev-goStumble/jev-goAn independent Go SDK for TypeSafe AI's Jev / System One API.SWswift-typesafeainame/swift-typesafeSwift 6.4 SDK for TypeSafe AI, following the Python SDK's 0.6.0 API.TYtypesafe_sdknshkrdotcom/typesafe_sdktypesafe_sdk is the Elixir SDK for TypeSafe AI and its first System One model, Jev.TYtypesafe_sdk_exvinnie357/typesafe_sdk_exTypesafe AI SDK in Elixir using ReqTYtypesafe-aiTwister915/typesafe-aitypesafe-ai brings TypeSafe's System One evaluation API into Rust as small, typed judgments that fit inside ordinary application code.TYtypesafe-ai-rsgilljon/typesafe-ai-rsAn independent Rust client for the TypeSafe AI System One API, maintained at gilljon/typesafe-ai-rs.TYtypesafe-rsAbdelStark/typesafe-rsEvaluate a state against named questions (noul, choice, score) and get one typed answer per question.TYtypesafe-sdkjoshmn/typesafe-sdkA Ruby client for the TypeSafe System One API.TYtypesafe-sdk-goTangerg/typesafe-sdk-goGo SDK for the TypeSafe AI API — typed questions in, probability distributions out.TYtypesafe-sdk-phpButochnikov/typesafe-sdk-phpPHP client for the TypeSafe AI System One API.TYtypesafeai-dotnet-sdksaibimajdi/typesafeai-dotnet-sdkCommunity .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.ZIzio-typesafe-aijamesward/zio-typesafe-aiA Scala 3 / ZIO library for TypeSafe AI's Jev / System One API — a "System One model" that answers typed, atomic questions about a piece of state instead of generating text.JEJev MCPjkudish/jev-mcpMCP server that uses TypeSafe Jev for claim verification, content screening, and semantic candidate ranking.RUruby_llm-typesafekieranklaassen/ruby_llm-typesafeTypeSafe structured-output provider for RubyLLM 2TYtypesafe-ai-railsGenieRobot/typesafe-ai-railsCommunity Rails integration for TypeSafe AI's System One API, built on the community typesafe-sdk Ruby gem.TYTypeSafe AI Playgroundmarkjaquith/typesafe-ai-playgroundRust CLI with TypeSafe Jev experiments for PHI detection, code-comment review, tone analysis, and classification.
Agents and guardrails
AGagent-routernidhi-singh02/agent-routerPre-release Herdr integration that filters eligible coding models by quota and policy before Jev ranks them.BLblinkellipsis-dev/blinkNavigates file and directory names with Jev-guided walkers to find codebase paths for a natural-language query.CACannyqkal/CannyEvidence ledger that challenges unsupported "done" claims from coding agents.COcommit-minerdevanshbatham/commit-minerClassifies Git diffs and commit messages into change types and candidate security-fix/CWE labels for inspection.FOforemanthruwire/foremanSoftware-factory supervisor that uses Jev to keep coding agents on task.ISis-maliciousluantak/is-maliciousScans source, configuration, build, and CI files with Jev, then reports suspicious behavior and implicated lines before the code is run.JEjev-codex-router0xNatoshi/jev-codex-routerPer-turn Codex model, reasoning, and speed-mode routing.JEjev-guardleepokai/jev-guardCross-agent tool-call risk scoring with allow, ask, and deny outcomes.JEjev-prefdoeixd/jev-prefLinter that has Jev check code changes against project preferences from `jev-pref.json` and feeds findings back to coding agents.JEjev-reviewdevagrawal09/jev-reviewStaged code-review workflow with a local dashboard.JEjev-review MCP pluginNiazMorshed2007/jev-reviewLocal-first continuous software-quality review for coding agents.JEjev-routergargpratyush/jev-routerChooses a model for each fresh Claude Code or Codex turn while wrapping the existing CLI.JEjevwireBrainwires/jevwireMCP tools, an embeddable decision library, and advisory or restrictive Claude Code hooks; judgments do not grant native permissions.OPopencode-jev-orchestratoraaronshaf/opencode-jev-orchestratorKeeps an OpenCode parent model fixed and uses Jev difficulty judgments to delegate harder turns to temporary subagents.PEperchlakeday-org/perchSemantic code linter that evaluates code units against configurable Jev questions.PIpi-jevy0usaf/pi-jevMeasured tool-call gate and general typed decision layer for the Pi coding agent.PIpi-wardenDevMortimer/pi-wardenPi extension that judges rule compliance, risky actions, stuck loops, and completion claims; enforcement depends on the hook and policy.SKskillboxkitze/skillboxSelf-hosted skill library with optional Jev relevance recommendations over an authorized catalog.SKskillrankerDicklesworthstone/skillrankerRust CLI that ranks agent skills against live session context and can abstain.SUsupercovsupercorp-ai/supercovScores source files so coding agents can prioritize code-quality work.WAwakegateshitianfang/wakegateExperimental gate where Jev decides whether a timer or incoming event is worth resuming a sleeping agent's LLM; code skips only when Jev is confident and always wakes on user messages, errors, and a skip limit.AIhermes-jevkeeltrace/hermes-jevIt is built for narrow, typed judgments that should not require the main generative model to improvise an answer: routing, ranking, verification, multi-question assessment, tool gating, and context-value decisions.AIJev Review Actionfatwang2/jev-review-actionConfigurable GitHub Action using Jev to review directory submissions and classify pull requests, with evidence links and template-generated comments.AIjev-codedevagrawal09/jev-codejev-code is a command-line toolkit that coding agents can delegate judgment-heavy work to.AIjev-gitAkashPriyadarshii/jev-gitSub-second Git pre-commit and pre-push reflex gate that screens staged diffs for secrets and destructive commands using TypeSafe Jev.AIjev-scoutAkashPriyadarshii/jev-scoutZero-hallucination open-source repo and crate scout powered by TypeSafe AI Jev System One scoring.AIjev-superpowersAkashPriyadarshii/jev-superpowersSystematic software development framework for AI coding agents upgraded with TypeSafe Jev System One typed decisions, zero-hallucination package vetting, and completion gates.AIJevSeekmorcoan/JevSeekDesktop and CLI coding agent that uses Jev to select the next tool from user intent and recorded execution results.AILeanestbaronunread/leanestLocal-first test selector that uses TypeSafe Jev semantic judgments to decide which tests are safe to skip for a given code change, dropping straight into CI via its bundled GitHub Action.
Routing and workflows
DUduckdb-jevcolliber/duckdb-jevDuckDB extension that exposes Jev judgments as SQL values with return types derived from the declared criteria.HAHA-JevAboveColin/HA-JevHome Assistant integration exposing typed answers as sensors, automation actions, and an Assist conversation agent.HOhono-jev-routeryusukebe/hono-jev-routerRoutes Hono HTTP requests by meaning.JEjev-curateAkashPriyadarshii/jev-curateStreaming filter and scorer for Parquet and JSONL datasets.JEjev-reviewerchoxos/jev-reviewerResearch-document extraction aid where Jev selects and verifies source lines for verbatim quotes; findings require human review and are not clinical decisions.JEjev-searchsuperagents-lab/jev-searchUses Jev to select search sources and rank Search1API results, returning source links and snippets.JEjev-traderjarrodwatts/jev-traderKuru/Monad trading experiment with optional Jev buy/sell decisions; defaults to a mock model and dry-runs without a private key, but configured execution can place real orders.JEjev-treereachjalil/jev-treeRecursive choice over taxonomies larger than Jev's direct option limit.JEjevlogsreachjalil/jevlogsOpenTelemetry log triage before more expensive analysis.JEjevqlkylemclaren/jevqlpsql-shaped client and Go/TypeScript/Python SDKs for vanilla Postgres where Jev makes Noul, Choice, and Score judgements about individual table rows after the plain SQL has run on the server, and the client applies the resulting filter, sort, or group.JEjevsqlEugeneBoondock/jevsqlSQL-like filtering, ranking, classification, and scoring with natural-language predicates.LLllama-index-jevWiktorB2004/llama-index-jevLlamaIndex reranker and selector using Jev Score and Choice answers, with configurable confidence handling.N8n8n-nodes-typesafe-jevn3ndor/n8n-nodes-typesafe-jevCommunity n8n node for asking multiple typed questions over workflow state.PGpg-jevrealZachi/pg-jevPostgreSQL extension for semantic questions over table rows.PGpg_typesafegiuliosmall/pg_typesafePre-alpha PostgreSQL C extension exposing Choice, Noul, Score, and batched judgments from SQL.TAtax-doc-classifierkyotofin/tax-doc-classifierClassifies text-bearing PDF pages into IRS form and page-kind candidates with a confidence gate; document triage, not tax advice, and scanned pages need OCR.TItiershiftiamvatsalpatel/tiershiftPolicy-bounded model routing for TypeScript and Python.TYtypesafe-jev-workflowGiesN/typesafe-jev-workflowLangGraph email-intent workflow using a typed Jev choice.HAha-conversation-jevluxus/ha-conversation-jevHome Assistant custom conversation agent: Jev (TypeSafe System One, jev-latest) classifies an utterance, then either calls a light service (v0 fast path) or hands off to the SpaceXAI Grok conversation agent.JEjev-starterhamakyo/jev-starterTyped, policy-driven decision workflows on top of TypeSafe AI Jev: confidence routing, fallbacks, evaluation, and RAG patterns for TypeScript apps.JEjev-seoAkashPriyadarshii/jev-seoZero-cost, agent-first SEO & Generative Engine Optimization (GEO) search radar CLI suite and MCP server powered by DuckDuckGo and TypeSafe Jev System One.HEhev rerankerhev/rerankerPython library using Jev Noul judgments to score candidate documents for query relevance, then sort or filter the results.NEneo4jevjexp/neo4jevNeo4j graph navigation demo that uses TypeSafe Jev to select relationships and check goals during beam search.JEJev-Tradeszadescoxp/Jev-TradesA Next.js dashboard for live crypto market data and TypeSafe-powered paper trading.PRPrismirfndi/prism-liquidity-agentA Solana liquidity agent with optional Jev shadow or advisory judgments. The integration is off by default and does not directly drive ENTER or EXIT; separate risk gates remain in control.
Browser and computer use
JEjev-browserjkudish/jev-browserBrowser-use experiment powered by Jev decisions.JEjev-browser-usewy-coliney/jev-browser-useCodex browser skill that uses Jev for navigation and target selection while Codex handles text entry and outcome verification.JEjev-ultrafastbrowser-use/jev-ultrafastBrowser agent where Jev selects an operation and compatible DOM target, and a separate LLM supplies typed text.JEjev-voice-browsermoritzkremb/jev-voice-browserMaps partial speech transcripts to browser intents and observed targets, with code deciding whether to act, wait, or ask.MOmobile-jevdroidrun/mobile-jevAndroid agent using Mobilerun observations and bounded Jev actions; execute mode controls a real device, while the published Uber demo stops before booking.TYtypesafe-computer-useawlevin/typesafe-computer-usemacOS computer-use experiment using OCR plus bounded Jev action selection.JEjev-browserYing-Kai-Liao/jev-browserBrowser automation where an LLM plans and Jev (Typesafe System One) decides. Library, CLI and MCP server.JEjev-egoromaluev/jev-egoTypeScript browser agent for ego lite — Jev Ultrafast's indexed action space, without Chrome, Playwright, or Browser Harness.CUjev-desktopyikangy873-gif/jev-desktopAdds a bounded Jev decision loop to Codex Computer Use: Jev selects among allowed actions and controls while Codex keeps scope, prepared text, permissions, recovery, and verification; the project does not claim a proven end-to-end speedup.
Context and compression
FAfast-dev-compactionleonaaardob/fast-dev-compactionCodex port that restores Jev-selected verbatim history around native session compaction.FAfast-jev-compactiontamaratran/fast-jev-compactionClaude Code plugin and library that score tool-call/result pairs for deletion or truncation while retaining selected text verbatim.PIpi-fast-jev-compactionjoelhooks/pi-fast-jev-compactionPi extension that prunes stale tool history and leaves summary compaction to Pi when pruning is insufficient.PIpi-jev-compactilkerulusoy/pi-jev-compactSelective, verbatim context compaction for Pi using Jev model.PIpi-jev-contextkevinpita/pi-jev-contextOpt-in Pi extension that filters older messages from model requests while preserving the original session history.WIwinnowGhalebDweikat/winnowJudges tool results before admitting them into Claude Code context.YOyoshicompozy/yoshiExperimental Claude Code/Codex proxy that uses Jev to prune request context while preserving tool-call protocol structure.
Games and demos
HEheist-oneAbdelStark/heist-oneBrowser stealth game where Jev judges guards while deterministic code owns the world.JEjev-droneRomanSlack/jev-droneSimulated MuJoCo quadrotor with Jev making slower tactical judgments from processed camera observations; deterministic code controls flight.JEjev-experimentsdabit3/jev-experimentsCollection of inspectable Jev demos, including scripted support conversations with typed intent, escalation, and suggested-response decisions.JEJevPilotstandardagents/jevpilotThree.js driving simulation where Jev chooses among candidate paths and speeds while local code handles vehicle dynamics and geometry.JEJevScapeSkyvern-AI/jevscapeRuneBench-based RuneScape harness that maps Jev choices to a bounded game-action catalog and records tick-level results.KIkillmyideamonteduro/killmyideaStartup-idea evaluator that chooses kill, fix, or ship.TStsai-scphyous/tsai-scOriginal StarCraft shareware controlled with recorded Jev action probabilities.TYtypesafe-mariofhshaik/typesafe-marioSuper Mario Bros. agent choosing actions from structured emulator state.TYtypesafe-snakesorrycc/typesafe-snakeSnake autoplayer with one typed decision per tick and code-generated legal moves.TYtypesafe-ai-playgroundBunsDev/typesafe-ai-playgroundA community playground for TypeSafe AI's Jev: edit classification experiments, compare A/B inputs, route conversations, extract document fields, inspect code-policy decisions, and explore games and simulations built around typed model outputs.
Media tools
JEjevmeterChetasLua/jevmeterScores every sentence in a video and renders the result as an overlay.JEJevthovencocktailpeanut/jevthovenSymbolic-music studio where Jev chooses plans, instruments, and bar patterns, and code renders editable music and MIDI.SLSlidePilotharshil1712/slidepilotExperimental Slidev controller that judges speech transcripts for slide completion, with deterministic checks and manual navigation.SPSponsor Skiptrungdq88/youtube-sponsor-detectionFinds sponsor reads in YouTube transcripts or transcribed audio while code owns timestamps and playback skipping.UNunclutterkitze/unclutterBrowser extension that uses Jev to identify page clutter and saves reusable, reversible hiding rules.VIVibe Check for XRafalWilinski/vibecheckChrome extension that scores draft posts and reply context before posting; optional media descriptions come from a separate vision model.UIUI Generator Instinct Jevjoevidev/ui-generator-instinct-jevA demo app where the user describes a UI case in free text and Jev (TypeSafe's System One model) answers typed Choice/Noul/Score questions to pick and configure a real shadcn/ui component or page block, never generating code or copy.
Research and reproductions
JEJev Visualhr98w/jev-visualEducational MLX/Qwen vision-language experiment sharing image context across candidate-scoring questions; its probabilities are not calibrated correctness estimates.JEJevlikevinnylarouge/jevlikeTrainable encoder and option-attention head for variable candidate sets, with separate visual game experiments.LILitJevzhengxuyu/litjevReproduction of Jev that turns any Qwen model into a fast decision model, serving the same `/v1/systemone` schema (Choice, Score, Noul) with no training and no generated answer text.JEjevmlxbnsd55/jevmlxJev-style parallel constrained decisions for MLX models on Apple Silicon.KEkevjaredpalmer/kevQwen2.5-0.5B adapter and decision head with training code, released weights, and parallel typed-question inference.NANanoJevTianyuCodings/NanoJevSmall parallel-decision model with dynamic candidates, a training pipeline, and recorded game comparisons that include shared code planning.OPopenjevzhihz/openjevLocal bilingual probability decisions from context, questions, and candidate answers.OPOpenJev (DiffusionGemma)razorback16/openjevIndependent Jev-compatible server over DiffusionGemma/vLLM; the documented setup requires custom vLLM patches.OPopenjev-sglangekzhang/openjev-sglangJev-compatible API endpoint backed by open models and prefill-only inference.OPopenvonsgenai-craft/openvonsOpen decision layer for finite options across text, images, and Japanese voice commands.PAparallelConstraintDecodingstephanj/parallelConstraintDecodingJava and llama.cpp experiments in parallel constrained decoding.REreflexkshetrajna12/reflexOpen-model decision engine with shared-state inference, isolated question branches, and a WebGPU demo; browser and Python configurations differ.SESemIfTheoLeeCJ/SemIfFormerly OpenJev: an independent study of typed option readout from frozen open models, with shared-prefix experiments and a WebGPU demo.SISimple Jevfeatherless-ai/simple-jevTransforms compatible open-model logits into typed decisions without a separately trained classifier head; model compatibility is constrained.VEVerdict-open-jevHeman10x-NGU/Verdict-open-jevModernBERT decision engine with calibrated uncertainty and a WebGPU playground.
Evaluation
JAJanusFirasSX914/JanusMeasures when to use Jev versus other models and routes accordingly.JEjev-behavior-studyRINNECODER/jev-behavior-studyIndependent synthetic-task study of Jev 1.13.0 framing sensitivity and failures, with raw responses and offline report checks.JEjev-benchmarksAbdelStark/jev-benchmarksReproducible evaluation for calibration, selective risk, and latency.JEjev-eval4esv/jev-evalIndependent Jev versus GPT-5.6 Terra comparison on three labeled classification tasks, reporting accuracy, calibration, latency, and cost.JEjev-eval-agentvinilana/jev-eval-agentCompares LLM tool selection with Jev routing in a personal-assistant harness containing 100 mocked tools.JEjev-korean-benchmarkmahlernim/jev-korean-benchmarkSmall Korean/English sample study with recorded responses, including medical-text questions; not a clinical validation.JEjev-search-rerank-evalzhuyansen/jev-search-rerank-evalChinese/English retrieval evaluation comparing Jev reranking with lexical, embedding, and fusion baselines, including judge-circularity analysis.JEjevcalabhixhek/jevcalFits and drift-checks confidence thresholds against labeled data.TYtypesafe-ai-benchmarkiammrduncan/typesafe-ai-benchmarkLLM gateway that mimics the System One output shape for comparison work.
Guides
Unofficial developer resource. Not affiliated with TypeSafe AI. These links point at third-party repositories. Read each repo before you depend on it.