Awesome Jev

Review queue

These repositories fell between the gates: Jev was confident enough not to drop them, but not confident enough to list them. The policy leaves them here rather than guessing. A human reading one repository for thirty seconds usually settles it.

If you can tell which way a row should go — or if you spot a wrong call anywhere in the index — open an issue with the Jev got it wrong template. Corrections become overrides in the data, and overrides feed the calibration set.

148 of 1445 judged repositories are awaiting review.

  • Use Jev (TypeSafe's System One model) as a calibrated reranker: one call, up to 30 documents, a probability per document. Apache-2.0.

    genuine0.97cat0.45subst1.74ApplicationsPython

    held because: category confidence 0.45 below 0.5

  • Sidecar: grade a cloudforge student writeup with TypeSafe Jev. Not part of the OSS product.

    genuine0.97cat0.46subst2.00ApplicationsPython

    held because: category confidence 0.46 below 0.5

  • No description provided.

    genuine0.97cat0.39subst1.81Agent and Dev ToolingTypeScript

    held because: category confidence 0.39 below 0.5

  • Verbatim Jev-scored context reduction for omp, over TypeSafe or OpenRouter

    genuine0.96cat0.43subst2.01Agent and Dev ToolingTypeScript

    held because: category confidence 0.43 below 0.5

  • Evidence-driven frontend QA built on Jev Ultrafast and Browser Harness, with a synthetic todo demo.

    genuine0.96cat0.30subst2.01Agent and Dev ToolingPython

    held because: category confidence 0.30 below 0.5

  • No description provided.

    genuine0.96cat0.47subst2.00ApplicationsTypeScript

    held because: category confidence 0.47 below 0.5

  • No description provided.

    genuine0.96cat0.45subst1.80Agent and Dev ToolingJavaScript

    held because: category confidence 0.45 below 0.5

  • macOS computer use driven by Jev (TypeSafe System One) as the decision maker

    genuine0.96cat0.48subst1.85ApplicationsGo

    held because: category confidence 0.48 below 0.5

  • A playground for experiments around Jev, TypeSafe's System One model.

    genuine0.96cat0.42subst2.00LearningRust

    held because: category confidence 0.42 below 0.5

  • jevil is in the details

    genuine0.96cat0.35subst1.22Agent and Dev ToolingJavaScript

    held because: category confidence 0.35 below 0.5

  • System 1 / System 2 reflex library: Jev triages real-time events, a reasoning model writes the workflows it can't yet handle.

    genuine0.95cat0.44subst2.02Agent and Dev ToolingPython

    held because: category confidence 0.44 below 0.5

  • TUI Sandbox for Typesafe Jev API

    genuine0.95cat0.35subst2.64LearningPython

    held because: category confidence 0.35 below 0.5

  • Use Jev (TypeSafe's System One model) as a calibrated reranker: one call, up to 30 documents, a probability per document. Apache-2.0.

    genuine0.95cat0.47subst1.76ApplicationsPython

    held because: category confidence 0.47 below 0.5

  • Typesafe.ai System One Model Jev navigating a Neo4j graph by using a classifier over neighbouring relationships

    genuine0.95cat0.49subst2.00LearningJupyter Notebook

    held because: category confidence 0.49 below 0.5

  • Ask typed questions of a codebase.

    genuine0.95cat0.46subst2.00SDKs and ClientsTypeScript

    held because: category confidence 0.46 below 0.5

  • No description provided.

    genuine0.95cat0.44subst1.90Agent and Dev ToolingGo

    held because: category confidence 0.44 below 0.5

  • No description provided.

    genuine0.95cat0.24subst1.04OtherJavaScript

    held because: category confidence 0.24 below 0.5

  • Typed judgments in, control flow out: two Jev calls route a support ticket to an agent, then pick its model tier and reasoning depth.

    genuine0.95cat0.42subst1.10LearningHTML

    held because: category confidence 0.42 below 0.5

  • Fast structured Android control loops with TypeSafe Jev and Mobile MCP

    genuine0.94cat0.45subst2.53ApplicationsPython

    held because: category confidence 0.45 below 0.5

  • Build versioned judgment functions on TypeSafe's Jev once, then call the same published version from your backend over HTTP and from coding agents over MCP. The vendor key stays on your machine.

    genuine0.94cat0.41subst2.08Agent and Dev ToolingTypeScript

    held because: category confidence 0.41 below 0.5

  • Bounded exploratory browser testing with Jev, deterministic assertions, and replayable evidence.

    genuine0.94cat0.33subst2.04SDKs and ClientsTypeScript

    held because: category confidence 0.33 below 0.5

  • Interactive TypeScript coding CLI powered by Jev typed decisions and constrained AST generation.

    genuine0.94cat0.35subst2.04ApplicationsTypeScript

    held because: category confidence 0.35 below 0.5

  • Neon Function proxy for the Neon AI Gateway with TypeSafe Jev routing.

    genuine0.94cat0.31subst1.83ApplicationsTypeScript

    held because: category confidence 0.31 below 0.5

  • Put your code comments on trial. Powered by Jev.

    genuine0.94cat0.49subst1.93SDKs and ClientsTypeScript

    held because: category confidence 0.49 below 0.5

  • Measure when to use Jev and other models on your data, then route accordingly.

    genuine0.93cat0.48subst2.65Agent and Dev ToolingPython

    held because: category confidence 0.48 below 0.5

  • A small, extensible decision-to-action harness for TypeSafe Jev

    genuine0.93cat0.49subst2.39Agent and Dev ToolingPython

    held because: category confidence 0.49 below 0.5

  • The intelligence layer for beads: typed, calibrated labels for your backlog

    genuine0.93cat0.49subst2.45Agent and Dev ToolingPython

    held because: category confidence 0.49 below 0.5

  • Helping JEV speak <3

    genuine0.93cat0.33subst1.03LearningTypeScript

    held because: category confidence 0.33 below 0.5

  • Bare-bones OpenRouter + TypeSafe Jev classifier: route legal prompts to Texas statute databases

    genuine0.93cat0.44subst1.15ApplicationsPython

    held because: category confidence 0.44 below 0.5

  • No description provided.

    genuine0.93cat0.99subst0.16LearningJava

    held because: substance 0.16 below 0.5

  • 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.

    genuine0.92cat0.44subst1.97Agent and Dev ToolingTypeScript

    held because: category confidence 0.44 below 0.5

  • Jev cannot generate a single note. Given a piano and the right questions, it improvises anyway.

    genuine0.92cat0.30subst1.42ApplicationsTypeScript

    held because: category confidence 0.30 below 0.5

  • No description provided.

    genuine0.92cat0.43subst1.93Agent and Dev ToolingPython

    held because: category confidence 0.43 below 0.5

  • Playful ticket triage that shows how TypeSafe Jev's typed, probability-backed decisions compare with prompting an LLM. Cost of Delay ranking, certainty-gated routing, bring-your-own-key LLM race.

    genuine0.92cat0.33subst1.70ApplicationsTypeScript

    held because: category confidence 0.33 below 0.5

  • Natural-language PostgreSQL predicates with TypeSafe Jev. A simple search playground with randomly generated databases.

    genuine0.92cat0.41subst1.95IntegrationsTypeScript

    held because: category confidence 0.41 below 0.5

  • Pre-install security gate for npm lifecycle scripts using TypeSafe System One.

    genuine0.92cat0.39subst2.09SDKs and ClientsJavaScript

    held because: category confidence 0.39 below 0.5

  • Neon Function proxy for the Neon AI Gateway with TypeSafe Jev routing.

    genuine0.92cat0.34subst1.79ApplicationsTypeScript

    held because: category confidence 0.34 below 0.5

  • TypeSafe Jev as a bounded judgement layer for computer use: verify, pick, classify over Cua Driver and phone-harness

    genuine0.91cat0.44subst2.00SDKs and ClientsPython

    held because: category confidence 0.44 below 0.5

  • I kept watching coding agents burn context on decisions that aren't hard - triage 400 tickets, tag 600 files, route to one of six teams. jev-mode moves those verdicts to a typed-judgment model. I A/B'd it: 78% fewer tokens, 16x less work-attributable input, accuracy 96.1% vs 93.7%. Python, no deps, MIT.

    genuine0.90cat0.43subst2.08Agent and Dev ToolingPython

    held because: category confidence 0.43 below 0.5

  • Jev (TypeSafe) 性能評価プロジェクト — 日本郵便 KEN_ALL をマスタに、AI SDK 経由の Jev が住所のあいまい一致にどこまで使えるかを検証

    genuine0.90cat0.43subst2.00ApplicationsTypeScript

    held because: category confidence 0.43 below 0.5

  • TypeSafe Jev (System One judgments) for Pi: zero-dependency jev_ask tool with question linting, model discovery, and budget caps

    genuine0.90cat0.41subst2.02Agent and Dev ToolingTypeScript

    held because: category confidence 0.41 below 0.5

  • Rubric-based eval harness cheap enough to run on every PR, powered by typesafe-ai/jev

    genuine0.90cat0.39subst2.01Research and EvalsTypeScript

    held because: category confidence 0.39 below 0.5

  • No description provided.

    genuine0.89cat0.36subst1.95IntegrationsTypeScript

    held because: category confidence 0.36 below 0.5

  • Verbatim, Jev-guided context compaction for Pi: drop stale tool calls and outputs, never summarize.

    genuine0.88cat0.41subst2.02Agent and Dev ToolingTypeScript

    held because: category confidence 0.41 below 0.5

  • Experimental multi-horizon BTC signal generator using TypeSafe Jev probabilities and Binance market data.

    genuine0.88cat0.47subst1.99Research and EvalsTypeScript

    held because: category confidence 0.47 below 0.5

  • No description provided.

    genuine0.88cat0.36subst2.00ApplicationsTypeScript

    held because: category confidence 0.36 below 0.5

  • Reversible context pruning for Pi, powered by TypeSafe Jev. Keep useful context without deleting session history.

    genuine0.88cat0.36subst2.25ApplicationsTypeScript

    held because: category confidence 0.36 below 0.5

  • Fake autoregressive language model powered by TypeSafe Jev

    genuine0.88cat0.49subst1.78ApplicationsPython

    held because: category confidence 0.49 below 0.5

  • custom linter with treesitter + jev

    genuine0.87cat0.49subst2.58Agent and Dev ToolingPython

    held because: category confidence 0.49 below 0.5

  • One DevTool for any Jev-powered application. Inspect every state, question, probability, decision and action. Replay decisions locally.

    genuine0.87cat0.46subst1.82SDKs and ClientsTypeScript

    held because: category confidence 0.46 below 0.5

  • A codebase scanner that answers is this obviously malicous before running

    genuine0.87cat0.49subst2.00ApplicationsTypeScript

    held because: category confidence 0.49 below 0.5

  • Context compaction for Hermes sessions that never summarises: every tool call is scored by TypeSafe's Jev model, stale calls are dropped, everything kept stays verbatim.

    genuine0.87cat0.49subst1.83Agent and Dev ToolingJavaScript

    held because: category confidence 0.49 below 0.5

  • An observable raw-character chat experiment powered entirely by TypeSafe Jev Choice

    genuine0.87cat0.39subst1.01LearningJavaScript

    held because: category confidence 0.39 below 0.5

  • Configurable GitHub submission review and PR classification with TypeSafe Jev. No text-generation model.

    genuine0.86cat0.41subst2.48Agent and Dev ToolingJavaScript

    held because: category confidence 0.41 below 0.5

  • Reliability and evaluation harness for financial agent workflows: agent harness + evaluation harness, release gates, evidence packs, DSPy programs vs typed tool-use (Jev). Reference prototype, synthetic and public benchmark data.

    genuine0.86cat0.44subst2.08Agent and Dev ToolingPython

    held because: category confidence 0.44 below 0.5

  • Jev powered OpenCode compaction

    genuine0.86cat0.47subst2.02IntegrationsTypeScript

    held because: category confidence 0.47 below 0.5

  • TypeSafe Jev-powered lead intake and live CRM judgment demo for Tribe CRM

    genuine0.86cat0.45subst1.81IntegrationsTypeScript

    held because: category confidence 0.45 below 0.5

  • High-speed recursive AI Elo tournament engine powered by Jev and Swiss matchmaking

    genuine0.84cat0.43subst1.88ApplicationsPython

    held because: category confidence 0.43 below 0.5

  • Hyprland/NixOS fork of typesafe-computer-use — grim + hyprctl + ydotool, TypeSafe Jev decisions

    genuine0.84cat0.32subst2.08IntegrationsPython

    held because: category confidence 0.32 below 0.5

  • Classify Git commit diffs and messages with Jev. Bug fixes, security fixes/CWEs, and change types.

    genuine0.83cat0.46subst2.00ApplicationsRust

    held because: category confidence 0.46 below 0.5

  • Jev-compatible decision engine over caller-provided models

    genuine0.81cat0.49subst2.00Agent and Dev ToolingPython

    held because: category confidence 0.49 below 0.5

  • No description provided.

    genuine0.81cat0.47subst1.94Research and EvalsJavaScript

    held because: category confidence 0.47 below 0.5

  • Find the AI agent runs that broke because their environment did: missing keys, tools, files, permissions, network, or context. Powered by TypeSafe Jev.

    genuine0.80cat0.43subst2.00Agent and Dev ToolingPython

    held because: category confidence 0.43 below 0.5

  • An asyncio daemon that tails journald or syslog and asks Jev, for each event, which subsystem it belongs to, how severe it is, and whether it's noise. Plain code then ignores the event, alerts on it, or escalates it to a human when confidence is low.

    genuine0.80cat0.99subst0.37Applications

    held because: substance 0.37 below 0.5

  • ACME live support-call scoring demo with TypeSafe AI, Effect, SQLite, React, Vite, and Turborepo

    genuine0.79cat0.44subst1.14ApplicationsTypeScript

    held because: category confidence 0.44 below 0.5

  • Documentation site for Syllago—the package manager for AI coding tool content

    genuine0.76cat0.45subst2.14OtherMDX

    held because: category confidence 0.45 below 0.5

  • Computer-use agent for a throwaway Linux desktop: AT-SPI accessibility-tree perception, TypeSafe Jev / GPT-5.6 planners, a11y-invoke or xdotool actions, Docker + noVNC, verdicts and screenshots exported per run.

    genuine0.76cat0.43subst1.74Agent and Dev ToolingPython

    held because: category confidence 0.43 below 0.5

  • Binary LLM-judge bench — compare Jev (TypeSafe System One) against a frontier LLM judge on speed, cost, and agreement with human labels.

    genuine0.76cat1.00subst0.09Research and Evals

    held because: substance 0.09 below 0.5

  • Open-source Jev log triage for OpenTelemetry. Score the signal before expensive LLM analysis.

    genuine0.74cat0.43subst2.38ApplicationsTypeScript

    held because: category confidence 0.43 below 0.5

  • A macOS computer-use agent with a dynamic, indexed action space. No screenshots, no coordinates. A macOS port of browser-use/jev-ultrafast.

    genuine0.74cat0.38subst1.70ApplicationsPython

    held because: category confidence 0.38 below 0.5

  • Sub-second Git pre-commit & pre-push semantic reflex gate powered by TypeSafe AI Jev

    genuine0.74cat0.37subst1.06Agent and Dev ToolingRust

    held because: category confidence 0.37 below 0.5

  • smarter than codegraph.

    genuine0.73cat0.49subst1.10ApplicationsTypeScript

    held because: category confidence 0.49 below 0.5

  • Reference prototype exploring agentic commerce: TypeSafe/Jev → Axiom → Argent → Silverscript on Kaspa.

    genuine0.72cat0.44subst1.60IntegrationsTypeScript

    held because: category confidence 0.44 below 0.5

  • Jev (TypeSafe System One) proof of concept: smart-home assistant demo

    genuine0.71cat0.49subst1.00LearningJavaScript

    held because: category confidence 0.49 below 0.5

  • config files

    genuine0.69cat0.49subst2.60Agent and Dev ToolingPython

    held because: category confidence 0.49 below 0.5

  • Open-source Jev-compatible MCP server for typed decisions (Choice/Score/Noul): local GLiNER backend with a model router, drop-in swappable with the TypeSafe System One API.

    genuine0.67cat0.96subst0.25Research and Evals

    held because: substance 0.25 below 0.5

  • Jev-style parallel constrained decisions for any MLX model on Apple Silicon. Typed, schema-valid JSON in one forward pass.

    genuine0.64cat0.44subst2.80Research and EvalsPython

    held because: category confidence 0.44 below 0.5

  • Prototyping against TypeSafe's Jev System One API: a one-ticket quickstart and a Bluesky Jetstream firehose demo.

    genuine0.63cat0.98subst0.37Learning

    held because: substance 0.37 below 0.5

  • Using the Jev model to compact agent tool logs. Supports integration with the Claude Agent SDK, the OpenAI Agent SDK, and more.

    genuine0.62cat0.64subst0.00Agent and Dev Tooling

    held because: substance 0.00 below 0.5

  • Jev-style parallel constrained decisions for any MLX model on Apple Silicon. Typed, schema-valid JSON in one forward pass.

    genuine0.61cat0.34subst2.76Research and EvalsPython

    held because: category confidence 0.34 below 0.5

  • Automated database migration safety reviewer powered by TypeSafe AI (Jev System One model)

    genuine0.61cat0.44subst1.82Agent and Dev ToolingTypeScript

    held because: category confidence 0.44 below 0.5

  • A Native R Client for JEV System One Decisions

    genuine0.61cat1.00subst0.19SDKs and ClientsR

    held because: substance 0.19 below 0.5

  • Using Jev to test how well it predicts financial markets(just like most llms as of september 2026, it doesnt do that good)

    genuine0.61cat0.93subst0.04Research and EvalsPython

    held because: substance 0.04 below 0.5

  • The typed social network. Agents post. Jev decides.

    genuine0.54cat0.84subst0.23Applications

    held because: substance 0.23 below 0.5

  • A System-1 model based memory retrieval reranked using caliberated decision space instead of embeddings

    genuine0.54cat0.74subst0.05Applications

    held because: substance 0.05 below 0.5

  • No description provided.

    genuine0.51cat0.34subst2.65Agent and Dev ToolingZig

    held because: category confidence 0.34 below 0.5

  • Jev (TypeSafe System One) × ASReview SYNERGY abstract screening demo — Choice/Noul vs gold labels

    genuine0.51cat0.50subst0.26Research and Evals

    held because: substance 0.26 below 0.5

  • TypeSafe Jev research + 5 product specs

    genuine0.50cat0.33subst0.15Other

    held because: category confidence 0.33 below 0.5

  • a Jev classification layer for the Pi coding agent

    genuine0.50cat0.73subst0.02Agent and Dev Tooling

    held because: substance 0.02 below 0.5

  • An evaluation of typesafe AI chess. As it turns out, the AI isn't doing really well even though chess is not a particularly open-ended game. Still, it's only a prototype and this probably wasn't optimzied for games.

    genuine0.49cat0.97subst0.97Research and EvalsPython

    held because: genuine 0.49 below 0.5

  • Pre-LLM catalog gate: TypeSafe/Jev scores tools & skills; keep only high-probability items

    genuine0.49cat0.97subst0.03Agent and Dev Tooling

    held because: genuine 0.49 below 0.5

  • Generate anything from your terminal

    genuine0.48cat0.64subst2.32SDKs and ClientsTypeScript

    held because: genuine 0.48 below 0.5

  • Local bilingual probability decisions from context, questions, and candidate answers. Independent research preview inspired by TypeSafe Jev.

    genuine0.48cat0.93subst2.01Research and EvalsPython

    held because: genuine 0.48 below 0.5

  • Code for Medium blog posts

    genuine0.48cat0.90subst1.07LearningJupyter Notebook

    held because: genuine 0.48 below 0.5

  • Trading project using JEV by TypeSafe AI

    genuine0.48cat0.99subst1.42ApplicationsPython

    held because: genuine 0.48 below 0.5

  • Apply OS — Personal career decision engine powered by TypeSafe AI (Jev). Ranks job postings, drafts applications, and automates the pipeline with typed, calibrated decisions. Agency-ready build spec included.

    genuine0.48cat0.92subst0.07Applications

    held because: genuine 0.48 below 0.5

  • A collection of what I use to keep things consitent

    genuine0.47cat0.38subst2.94SDKs and ClientsTypeScript

    held because: genuine 0.47 below 0.5

  • No description provided.

    genuine0.47cat0.99subst2.01LearningHTML

    held because: genuine 0.47 below 0.5

  • Agent plugin: TypeSafe Jev browser automation skills for Grok Bot / Cursor

    genuine0.46cat0.99subst1.14Agent and Dev Tooling

    held because: genuine 0.46 below 0.5

  • Judge agent steps — ok / retry / escalate / stop via TypeSafe Jev

    genuine0.46cat0.81subst0.77Agent and Dev ToolingTypeScript

    held because: genuine 0.46 below 0.5

  • PulseLane — clinic triage decisions via TypeSafe Jev

    genuine0.46cat1.00subst0.91ApplicationsTypeScript

    held because: genuine 0.46 below 0.5

  • No description provided.

    genuine0.45cat1.00subst2.06Agent and Dev ToolingPython

    held because: genuine 0.45 below 0.5

  • No description provided.

    genuine0.45cat0.79subst1.68ApplicationsJavaScript

    held because: genuine 0.45 below 0.5

  • Train lightweight language backbones for typed decisions and candidate probabilities. CE/Brier training, evaluation, and an offline end-to-end demo.

    genuine0.43cat0.99subst2.77Research and EvalsPython

    held because: genuine 0.43 below 0.5

  • Workload-aware Pareto frontiers and dynamic model selection for agentic systems

    genuine0.43cat0.91subst2.45Research and EvalsPython

    held because: genuine 0.43 below 0.5

  • No description provided.

    genuine0.43cat0.60subst1.21LearningPython

    held because: genuine 0.43 below 0.5

  • Like Claude Plays Pokemon, but with Jev

    genuine0.43cat0.86subst1.15Games and SimulationPython

    held because: genuine 0.43 below 0.5

  • Public repo containing generated synthetic data to test jev from typesafe

    genuine0.43cat0.74subst0.09Research and EvalsPython

    held because: genuine 0.43 below 0.5

  • An educational Jev-like visual inference experiment on Apple Silicon: shared context, direct candidate scoring, and local visual demos.

    genuine0.42cat0.85subst1.95Research and EvalsPython

    held because: genuine 0.42 below 0.5

  • Agent budget brake — continue / downgrade_model / stop via TypeSafe Jev

    genuine0.42cat0.44subst0.82ApplicationsTypeScript

    held because: genuine 0.42 below 0.5

  • Self-correcting audiobook TTS generator (CoreWeave Hacks) — TypeSafe tagging + Google TTS/STT + ARIA/Weave eval-correction loop

    genuine0.42cat0.93subst0.62ApplicationsPython

    held because: genuine 0.42 below 0.5

  • Passwortgeschützte Jev-Review-Auswertung; ausschließlich verschlüsselte HTML-Datei

    genuine0.42cat0.30subst0.10Research and EvalsHTML

    held because: genuine 0.42 below 0.5

  • No description provided.

    genuine0.41cat0.75subst2.07Agent and Dev ToolingGo

    held because: genuine 0.41 below 0.5

  • The open-source email platform — self-host on your own AWS SES, or use the cloud. Resend-compatible API.

    genuine0.40cat0.88subst2.35ApplicationsTypeScript

    held because: genuine 0.40 below 0.5

  • Self-hosted, versioned skills library for AI agents. MCP, scoped clients, and optional Jev recommendations.

    genuine0.40cat0.99subst2.04Agent and Dev ToolingTypeScript

    held because: genuine 0.40 below 0.5

  • Typed judgment layer for coding agents — gates from PRD to ship. Jev-ready, provider-agnostic.

    genuine0.40cat1.00subst1.66Agent and Dev ToolingTypeScript

    held because: genuine 0.40 below 0.5

  • No description provided.

    genuine0.39cat0.97subst1.68Research and EvalsHTML

    held because: genuine 0.39 below 0.5

  • Autonomous prospecting for freelancers.

    genuine0.39cat0.72subst1.10ApplicationsHTML

    held because: genuine 0.39 below 0.5

  • Match user goals to MCP catalog (two-stage) via TypeSafe Jev

    genuine0.39cat0.58subst0.80ApplicationsTypeScript

    held because: genuine 0.39 below 0.5

  • No description provided.

    genuine0.39cat0.93subst0.18LearningTypeScript

    held because: genuine 0.39 below 0.5

  • Stops AI coding agents from claiming work is done without evidence. Deterministic hooks decide, TypeSafe's Jev advises. Append-only ledger, zero runtime dependencies.

    genuine0.38cat1.00subst2.39Agent and Dev ToolingTypeScript

    held because: genuine 0.38 below 0.5

  • Route tasks to research/code/browser/support/writer agents via TypeSafe Jev

    genuine0.38cat0.80subst0.88Agent and Dev ToolingTypeScript

    held because: genuine 0.38 below 0.5

  • An AI coding agent that reads your code, edits files, runs commands, and reasons through bugs — across a terminal REPL, a full-screen TUI, and a browser UI, while you keep your hand on every permission.

    genuine0.37cat0.81subst2.70Agent and Dev ToolingTypeScript

    held because: genuine 0.37 below 0.5

  • High-performance LLM proxy: one OpenAI-style API over 20+ providers, with multi-tenant keys, full request logging, and usage analytics. Built in Rust.

    genuine0.37cat0.50subst2.51IntegrationsRust

    held because: genuine 0.37 below 0.5

  • No description provided.

    genuine0.37cat0.80subst2.03ApplicationsPython

    held because: genuine 0.37 below 0.5

  • Verbatim, decision-based context compaction for Prime Agent — instead of summaries, stale tool calls are scored and dropped; everything kept stays byte-for-byte intact.

    genuine0.37cat0.98subst0.56Agent and Dev Tooling

    held because: genuine 0.37 below 0.5

  • 3D chess powered by TypeSafe AI (Jev). AI vs AI by default, or play either side. Multiple difficulty levels.

    genuine0.37cat0.99subst0.79Games and SimulationTypeScript

    held because: genuine 0.37 below 0.5

  • A polished OpenAI + TypeSafe Jev terminal interface for answers with transparent decision reports

    genuine0.37cat0.66subst0.00SDKs and Clients

    held because: genuine 0.37 below 0.5

  • typesafe.ai model jev finance benchmark

    genuine0.36cat0.96subst0.00Research and Evals

    held because: genuine 0.36 below 0.5

  • mini-Jev: what a Jev-style typed-decision interface looks like on a frozen Qwen3-4B — read the option letter's logits instead of generating JSON. Preregistered experiment, results, teaching bench.

    genuine0.35cat1.00subst1.92Research and EvalsPython

    held because: genuine 0.35 below 0.5

  • 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.

    genuine0.35cat0.99subst2.11Research and EvalsPython

    held because: genuine 0.35 below 0.5

  • U-NSGA-III multi-objective evolutionary optimization for .NET (Seada & Deb) — ZDT/DTLZ, IGD oracle vs pymoo

    genuine0.35cat0.76subst2.75OtherC#

    held because: genuine 0.35 below 0.5

  • Local-first control layer for Pi. TypeSafe routes tasks, manages cache-aware model leases, gates bounded specialist consultations, protects context, verifies outcomes, and logs decisions for future training.

    genuine0.35cat0.99subst1.73Agent and Dev ToolingTypeScript

    held because: genuine 0.35 below 0.5

  • Testing Jev ai model

    genuine0.35cat0.38subst0.67Research and Evals

    held because: genuine 0.35 below 0.5

  • LaneBreak — support ticket priority+routing via TypeSafe Jev

    genuine0.35cat0.99subst0.52ApplicationsTypeScript

    held because: genuine 0.35 below 0.5

  • Model router for pi with Jev

    genuine0.35cat0.92subst0.00Agent and Dev Tooling

    held because: genuine 0.35 below 0.5

  • Grade page sections for clarity, writing and on-page SEO. WXT + TypeSafe AI Jev.

    genuine0.34cat0.96subst1.93ApplicationsTypeScript

    held because: genuine 0.34 below 0.5

  • NPCs of River Oaks Houston, Texas using Jev to power NPCs

    genuine0.34cat0.53subst0.03Games and Simulation

    held because: genuine 0.34 below 0.5

  • No description provided.

    genuine0.34cat0.50subst0.03SDKs and ClientsJavaScript

    held because: genuine 0.34 below 0.5

  • Typed-decision benchmark from PadFlow (land development SaaS): schemas, anonymized labeled rows, and a runner for confidence-calibrated models like TypeSafe Jev.

    genuine0.33cat1.00subst1.39Research and EvalsPython

    held because: genuine 0.33 below 0.5

  • HireSignal — resume first-pass fit+interview via TypeSafe Jev

    genuine0.33cat1.00subst0.47ApplicationsTypeScript

    held because: genuine 0.33 below 0.5

  • A Jev-inspired decision interface for existing LLMs. Explicit choices, scores, calibration, and review thresholds.

    genuine0.32cat0.29subst2.05IntegrationsHTML

    held because: genuine 0.32 below 0.5

  • Jev from Typesafe.ai + vphone-cli

    genuine0.32cat0.28subst2.21ApplicationsSwift

    held because: genuine 0.32 below 0.5

  • A test project based on Jev AI, the goal is to build a search function for a blog/article website that has 100s of articles to search from, So the user can actually use the search as chat to question anything and find related answers/articles

    genuine0.31cat0.99subst0.36ApplicationsJavaScript

    held because: genuine 0.31 below 0.5

  • Security scanning skill for AI coding agents. CVE vulnerability detection, OWASP Top 10 code analysis, and dependency audits. Follows agentskills.io specification.

    genuine0.30cat1.00subst2.93Agent and Dev ToolingPython

    held because: genuine 0.30 below 0.5

  • AI API server for common use cases — supports multiple models and providers. Run locally with Ollama or LM Studio, or in the cloud via OpenRouter, OpenAI, Anthropic, or Google.

    genuine0.30cat0.16subst2.21IntegrationsHTML

    held because: genuine 0.30 below 0.5

  • No description provided.

    genuine0.30cat0.70subst1.31ApplicationsPython

    held because: genuine 0.30 below 0.5

  • AI powered spam detection

    genuine0.30cat0.98subst0.48ApplicationsTypeScript

    held because: genuine 0.30 below 0.5