cobanov/awesome-jev

A curated, source-backed list of projects built with Jev, TypeSafe AI's System One model for typed decisions.

What it solves

Jev is a model designed for fast, typed, probabilistic decisions—such as classification, routing, scoring, ranking, verification, and guardrails—rather than free-form text generation. This list collects projects that build on or integrate with Jev, helping developers find ready-made tools, SDKs, and frameworks to incorporate these decision capabilities into their own applications.

How it works

The list is curated and source-backed, favoring projects with public code, concrete Jev usage, clear limitations, and reproducible evidence. It organizes entries into categories like official resources, provider integrations, framework integrations, SDKs and developer tools, agents and guardrails, context and compaction, browser and computer use, routing and workflows, games and robotics, media and creative tools, open reproductions, evaluation and calibration, and guides. Each entry typically includes a link to the repository and a short description of what it does. The README also warns that schema-valid output is not the same as a correct decision, and advises validating on your own data and keeping high-impact actions behind deterministic checks.

Who it’s for

Developers and teams interested in using Jev (or similar System One decision models) in their applications, who want to see real integration examples, official SDKs, provider gateways, framework adapters, and community-built tools across many programming languages and ecosystems.

Highlights

  • Catalog of 155 source-reviewed integrations, projects, and studies
  • Official resources: TypeSafe AI docs, JS/Python SDKs, system-one adapter, skills, model aliases, confidence guidance, and known limitations
  • Provider integrations: Cloudflare AI, Netlify AI Gateway, OpenRouter, Vercel AI Gateway
  • Framework integrations: LangChain, LangChain.js, LiteLLM, Pydantic AI, Rig, TanStack AI, Effect, BAML, Ax, Composio, and more
  • Community developer tools: hunch (Ruby/TS/Python), jev-mcp servers, jegrep/jgrep semantic grep, jev.nvim, pytest-jev, many language SDKs (Go, Java, Elixir, Rust, Kotlin, etc.)
  • Agent-focused tools: coding-agent routers, guardrails, malicious-code scanners, evidence-ledger systems
  • Open reproductions and research, plus evaluation and calibration guides

Related

  • Project
    ruvnet/Bot-Generator-Bot

    Bot Generator Bot turns a bot idea into a runnable agent by compiling a typed specification into a deterministic manifest, then running it via CLI or MCP with operator-configured model access. It includes a 39-prompt catalog, strict output validation, and bounded agent execution.

  • Project
    aronprins/paperclip-desktop

    An unofficial Electron wrapper for Paperclip that turns the AI agent orchestration platform into a one-click desktop application for easier installation and launch.

  • Project
    greenfield-inc/Pane

    Pane is a terminal-based workspace manager that automates git worktrees and process isolation to help developers run and coordinate multiple AI coding agents in parallel.

  • Project
    ViktorAxelsen/MemSkill

    A framework for long-horizon agents to learn and evolve data-driven memory skills, replacing static memory operations with a self-refining meta-memory system.