mcp_jev

# mcp_jev

Local MCP server that runs **TypeSafe Jev** (System One) **packs** — typed **Choice / Noul / Score** judgments, not chat. When no pack fits, **`run_questions`** is the typed custom path (still System One, still not an essay).

Anyone runs it **on their own PC** with **their own** TypeSafe API key. This repo does not host Jev, proxy your key, or invent a free-form `ask_jev` tool.

**Configure the TypeSafe key once** during MCP install (`~/.mcp_jev/.env`). Any agent that attaches this MCP reuses it. Do **not** paste the key into host `mcp.json` / TOML.

**Docs (source of truth):** this repo — [https://github.com/pedroknigge/mcp_jev](https://github.com/pedroknigge/mcp_jev) · TypeSafe API: [docs.typesafe.ai](https://docs.typesafe.ai) · [llms.txt](https://docs.typesafe.ai/llms.txt)

Skill: [skills/mcp_jev/SKILL.md](skills/mcp_jev/SKILL.md) · Custom judgments: [docs/CUSTOM_JUDGMENTS.md](docs/CUSTOM_JUDGMENTS.md) · Blind dogfood: [docs/DOGFOOD.md](docs/DOGFOOD.md) · Product backlog: [docs/BACKLOG.md](docs/BACKLOG.md) · Host deep dive: [docs/INSTALL_AGENTS.md](docs/INSTALL_AGENTS.md) · Releases: [docs/RELEASES.md](docs/RELEASES.md)

- [Why mcp_jev](#why-mcp_jev)
- [Install (happy path)](#install-happy-path)
- [First-run](#first-run)
- [Verify](#verify)
- [Recipes](#recipes)
- [Blind-test protocol](#blind-test-protocol)
- [Update](#update)
- [Skill stays in sync](#skill-stays-in-sync)
- [Say this to your agent](#say-this-to-your-agent)
- [Install for agents & IDEs](#install-for-agents--ides)
- [Tools](#tools-closed-catalog)
- [Troubleshooting](#troubleshooting)

## Why mcp_jev

mcp_jev is the **judgment layer** for agents and harnesses. Jev is fast at **typed decisions** (one Choice, a few Nouls, a Score) over a short named state. It is not a chat model and not an exe