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

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 serverbetaMIT
View on GitHub ↗

Overview

decisionkit gives coding agents, scripts and CI fast closed-form decisions: ask a question about text as yes/no, pick one, rate on a scale or rank candidates, and get probabilities rather than paragraphs. It is provider-neutral, supporting Jev via TypeSafe or via OpenRouter, the OpenAI Decisions API, and Claude through an existing Claude Code login, where probabilities are self-reported and calls are slower.

It can be used as a CLI, a Claude Code plugin, or an MCP server for Codex, Cursor, VS Code, Gemini CLI, Claude Desktop, Windsurf and Zed. The README shows examples such as classifying an issue, catching a breaking API change and ranking the files behind a bug, plus a recording where Claude hands triage of three issues to Jev and flags the one it is unsure about. The CLI reports provider, latency and cost on stderr.

What the decision model decides

TypeQuestionAnswers
ChoiceWhat kind of issue is this?bug · feature · refactor · docs
Yes / noDoes this change break the public API?(inferred)yes · no

How it works

  1. Choose a provider via DECISIONKIT_PROVIDER
  2. Pipe text to decisionkit choice/noul/score/rank
  3. Read the answer and probability on stdout
  4. Use the result in scripts, CI or agent tool calls

Strengths

  • Works with several decision providers
  • Integrates with many agents through MCP
  • Reports cost and latency per call

Caveats

  • Claude provider gives self-reported probabilities
  • Requires Node 20+

Getting started

npm i -g @adxtya21/decisionkit, set a provider, then echo "..." | decisionkit choice "What kind of issue is this?" bug feature refactor docs.

At a glance

Type
MCP server
Status
beta
Language
TypeScript
Runtime
CLI, MCP server, Claude Code plugin
Has code
yes
Has tests
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • Fast, cheap, probability-scored decisions for coding agents, scripts and CI. CLI + MCP server + Claude Code plugin for the OpenAI Decisions API and Jev (TypeSafe AI).

    hellogumbo/awesome-jev

Frequently asked questions

Do I need a decision API key?

Not with the Claude provider, which uses your Claude Code login; Jev and OpenAI need keys.

Which agents can use it?

Claude Code as a plugin, and Codex, Cursor, VS Code, Gemini CLI, Claude Desktop, Windsurf and Zed via MCP.

What question types exist?

Yes/no, choice, score and rank.

Related projects

  • jev-flash-router

    MCP server exposing Jev Noul, Choice and Score via OpenRouter so coding agents can make quick decisions without reasoning tokens.

    MCP serverbetaTypeScript
  • Awesome Jev Skills

    Installable agent skills for triage, routing, code review and document work with Jev, plus 108 scenarios and recorded examples.

    ArticlebetaPython
  • jev-code

    Gives Claude Code, Codex, Pi and OpenCode five typed Jev tools (classify, check, score, rank, ask) plus a skill on when to use them.

    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
  • JevX

    CLI and MCP tool that finds hard-coded judgment calls in your code, scores how well Jev fits, and replaces strong fits with a fallback.

    ProjectbetaTypeScript
  • mcp-server-jev

    Community MCP server giving Codex, Claude and other clients two tools for typed Jev decisions: evaluate and classify.

    MCP serverbetaTypeScript