Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

typesafe-jev-mcp

MCP server exposing Jev as a single evaluate tool for Claude Code, Codex, OpenCode, Antigravity or Cursor.

MCP serverbetaMIT
View on GitHub ↗

Overview

typesafe-jev-mcp is an MIT-licensed MCP server with one tool, evaluate: you pass state and typed questions and get noul, choice or score answers with probabilities. It is distributed as a crate, an npm package and release binaries, with install scripts for macOS, Linux and Windows that configure supported agents (Claude Code, Codex, OpenCode, Antigravity or Cursor) and prompt for a key.

The README points to TypeSafe's live docs and advises pinning a versioned model such as jev-1.13.0 once thresholds are tuned. It is not affiliated with TypeSafe and has CI. The example response shows a noul probability and a choice with per-option probabilities and confidence, which agents can compare against their own thresholds.

What the decision model decides

TypeQuestionAnswers
Yes / noDoes this convey urgency?yes · no
ChoiceWhich team should handle this?billing · technical

How it works

  1. Run the install script and paste a TypeSafe key.
  2. Restart the agent.
  3. The agent calls evaluate with state and questions.
  4. It acts on returned probabilities.

Strengths

  • One-command setup across several agents.
  • Multiple distribution channels (crate, npm, binaries).
  • Advice to pin model versions after tuning.

Caveats

  • Requires a TypeSafe key.
  • Unofficial.

Getting started

curl -fsSL https://raw.githubusercontent.com/anasbekheit/typesafe-jev-mcp/main/scripts/install.sh | sh

At a glance

Type
MCP server
Status
beta
Language
Rust
Runtime
MCP server
Has code
yes
Has tests
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • This repository is an MCP server for TypeSafe Jev that provides a single evaluate tool taking state plus typed questions and returning noul, choice, or score answers with probabilities.

    logicrw/awesome-jev-projects

  • MCP server exposing TypeSafe's Jev model as a typed evaluate tool.

    hellogumbo/awesome-jev

Frequently asked questions

What tool does it provide?

A single evaluate tool taking state and typed questions.

Which agents are supported?

Claude Code, Codex, OpenCode, Antigravity and Cursor.

Should I pin the model?

Yes, the README recommends pinning a versioned model after tuning thresholds.

Related projects

  • Building with Jev★ 155

    Agent skill for writing and improving programs that call Jev: question design, state structure, thresholds and diagnosing wrong answers.

    Agent skillproduction
  • askjev

    Unofficial hosted or local Jev MCP server giving agents a quick second opinion: it picks yes/no, scale or choice and returns probabilities.

    MCP serverbetaTypeScript
  • jev-mcp

    Python FastMCP server wrapping Jev's noul, choice and score judgments, with configs for many coding agents and an OpenCode Zen route.

    MCP serverbetaPython
  • jev-mcp-server

    PyPI MCP server for Jev with choice, score, noul, compare, verify and batch classify tools plus a one-command client installer.

    MCP serverbetaPython
  • jev-x-kit

    MCP server, CLI and Claude Code skill with Choice, Score and Noul primitives and a confidence gatekeeper, hosted or offline.

    MCP serverbetaTypeScript
  • prompt2jev

    Agent skill and CLI that converts plain requirements, LLM prompts or prompt-calling code into a typed Jev request and runnable script.

    Agent skillbetaPython