jev-agent-toolkit
Portable Agent Skill teaching coding agents to use Jev as a bounded judgment layer, plus an optional MCP bridge.
Overview
jev-agent-toolkit is a portable Agent Skill, with one canonical SKILL.md and no per-host forks, that teaches coding agents to build with Jev as a bounded judgment layer. It works in Claude Code, OpenAI Codex, Cursor and other Agent Skills-compatible hosts, and covers Jev concepts such as the decision gate along with capability-driven workflows for code, browser research, Blender and Unreal Engine.
An optional MCP bridge, published on npm as jev-agent-toolkit-mcp, exposes Jev to hosts over MCP. The README is unusually clear about status: the skill is available from GitHub, the bridge is implemented and tested, the npm install and handshake were verified on 20 September 2026, CI and protocol tests run against a local mock upstream, and manual live checks passed on direct HTTP, both SDKs, the bridge, and each host's MCP integration. It stresses these were one-off manual runs with no load testing, stability testing or evidence of general semantic accuracy. It is MIT licensed.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Other | Teaches agents to frame bounded Choice, Score and Noul gates in their own workflows(inferred) | — |
How it works
- Install the skill from GitHub into the agent host
- Agent reads SKILL.md and identifies decisions suited to Jev
- Optionally call Jev through the MCP bridge
- Agent acts on typed answers within its workflow
Strengths
- One skill works across Claude Code, Codex and Cursor
- Transparent verification record
- MCP bridge available via npx
Caveats
- No automated live tests; manual checks only
- No evidence of general semantic accuracy, per the README
Getting started
Install the skill from the GitHub repo; for MCP, run `npx -y jev-agent-toolkit-mcp` with TYPESAFE_API_KEY.At a glance
- Type
- MCP server
- Status
- beta
- Language
- TypeScript
- Runtime
- Agent Skill, MCP server
- Has code
- yes
- Has tests
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Jev-first portable Agent Skill and optional MCP bridge for Claude Code, Codex, Cursor and compatible agents.
logicrw/awesome-jev-projects
Frequently asked questions
What does the toolkit provide?
An Agent Skill that teaches agents to use Jev for bounded judgments, plus an optional MCP bridge.
Which hosts are supported?
Claude Code, Codex, Cursor and other Agent Skills-compatible hosts.
How well tested is it?
CI uses a mock upstream; live checks were manual one-off runs.
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 Independent agent skill that makes a coding agent design, build, test and tune a bounded Jev decision layer with a contract first.
Agent skillbetaClaude Code and Codex skills plus a study course for building with Jev, including Jevaluate, which rates how well projects use Jev.
ProjectbetaCommunity Jev skill for Claude Code and Codex: a guide, a dependency-free Python CLI and 5 measured question sets.
Agent skillbetaPythonAgent skill that finds fuzzy semantic judgments in a codebase and redesigns them as small Jev Choice, Score or Noul primitives.
Agent skillbetaAgent skill and CLI that chains Jev question design: find decision calls, check fit, draft, lint, calibrate thresholds and plan a cascade.
Agent skillprototypePython