jev-mcp (CodeIA-Academy)

Dependency-free Python MCP server (Spanish docs) exposing Jev to coding agents through ask_jev and list_jev_models.

MCP serverbetaMIT

Overview

This jev-mcp is a minimal MCP server written in standard-library Python that exposes TypeSafe's Jev to agents such as Claude Code, Codex, Cursor, Hermes and agents launched from Orca. It provides two tools: ask_jev, which sends a state and typed questions (choice, score, noul) to the Jev API, and list_jev_models. The documentation is in Spanish.

The README clarifies that TypeSafe documents an API and SDKs but no official MCP server, so this repository bridges the two. A run.sh wrapper reads the key from a gitignored .env file, because some MCP clients strip environment variables when starting servers, which makes third-party servers start fine but fail on every call. A smoke test performs a handshake, lists tools and makes one real call.

What the decision model decides

Type	Question	Answers
Choice	Pick one option from a list	—
Score	Place the state on an ordered scale	—
Yes / no	Probability that a statement is true	yes · no

How it works

Clone the repo and put the API key in .env
Run smoke_test.sh to verify a real call
Register run.sh with the agent as an MCP server
Agents call ask_jev with state and typed questions

Strengths

No dependencies beyond the Python standard library
Avoids the common missing-environment-variable failure
Smoke test included

Caveats

Documentation only in Spanish
Thin wrapper with two tools
Requires a TypeSafe API key

Getting started

git clone https://github.com/CodeIA-Academy/jev-mcp.git && cd jev-mcp && cp .env.example .env (add key) && bash smoke_test.sh, then claude mcp add jev --scope user -- /path/to/jev-mcp/run.sh

At a glance

Type: MCP server
Status: beta
Language: Python
Runtime: MCP server (stdio)
Has code: yes
Has tests: yes
License: MIT

Use cases and integrations

Infra / SDKs / Integrations537

MCP servers and agent skills334

How curated lists describe it

MCP local que expone Jev (TypeSafe) como herramienta para Claude Code, Codex, Hermes y cualquier agente: ask\_jev y list\_jev\_models, sin dependencias
logicrw/awesome-jev-projects
MCP local que expone Jev (TypeSafe) como herramienta para Claude Code, Codex, Hermes y cualquier agente: ask_jev y list_jev_models, sin dependencias.
hellogumbo/awesome-jev

Frequently asked questions

What tools does it expose?

ask_jev for typed questions and list_jev_models to list available models.

Why is there a run.sh?

It reads the key from .env so the server works even when an MCP client filters environment variables.

Does TypeSafe publish an official MCP server?

According to the README, no; TypeSafe publishes an API, SDKs and an agent skill.

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
decisionkit
CLI, MCP server and Claude Code plugin for closed-form yes/no, choice, score and rank decisions via Jev, OpenAI Decisions or Claude.
MCP serverbetaTypeScript
jev-agent-toolkit
Portable Agent Skill teaching coding agents to use Jev as a bounded judgment layer, plus an optional MCP bridge.
MCP serverbetaTypeScript
jev-mcp
MCP server giving any agent Jev's classify, score, yes/no check and allow/confirm/block gating of risky tool calls.
MCP serverbetaJavaScript
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