jev-codex-plugin
Codex plugin that consults Jev to choose tools, models and skills, diagnose failed commands and check completion claims.
Overview
Jev Workflows for Codex is an open-source plugin that lets a Codex session consult TypeSafe Jev about tools, models, tasks, skills, context, strategies and custom decisions. It also diagnoses failed commands and reviews completion claims against evidence. Version 0.4.0 ships an MCP plugin with eight tools, twelve lifecycle adapters and three standalone skills that work without MCP.
It is for Codex users who want typed decisions inside their agent loop without writing integration code. The plugin installs from a community Git marketplace; automatic consultation is off by default until the user enables it. Standalone skill ZIPs (decision classification, failure diagnosis, completion evidence review) bundle a prebuilt Node CLI that defaults to a local preview and only calls Jev when asked to evaluate. Abstentions or unavailable results never stop ordinary work.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Which tool, model, skill or strategy should Codex use next?(inferred) | — |
| Choice | What is the likely cause of this failed command?(inferred) | — |
| Yes / no | Does the evidence support the claim that the task is complete?(inferred) | — |
How it works
- Install the plugin from the marketplace and set TYPESAFE_API_KEY
- Ask Codex to check Jev status and enable automation if wanted
- Lifecycle hooks or MCP tools send state and typed questions to Jev
- Use the returned choice or verdict; abstention or errors fall back to normal work
Strengths
- Prebuilt entry points, no npm install needed
- Automation is opt-in and fails open
- Standalone skills usable without MCP
Caveats
- Requires Node.js 22+ and your own TypeSafe key; API charges apply
- Not listed in the official OpenAI plugin directory
- Hooks must be reviewed and trusted on install
Getting started
`codex plugin marketplace add integrate-your-mind/jev-codex-plugin` then `codex plugin add jev-workflows@jev-workflows`.At a glance
- Type
- Plugin
- Status
- beta
- Language
- TypeScript
- Runtime
- Codex plugin with MCP server and Node CLI skills
- Has code
- yes
- Has tests
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Open-source Codex plugin for TypeSafe Jev decision consultation, failure diagnosis, and evidence-based completion review
logicrw/awesome-jev-projectsOpen-source Codex plugin for TypeSafe Jev decision consultation, failure diagnosis, and evidence-based completion review.
hellogumbo/awesome-jev
Frequently asked questions
What does the plugin ask Jev?
Typed questions about tools, models, skills, strategies and custom decisions, plus failure diagnosis and completion review.
Is automation on by default?
No. Fresh installs keep automatic consultation disabled until you enable it.
Does it need a TypeSafe key?
Yes, TYPESAFE_API_KEY must be set in the Codex host environment; TypeSafe pricing and rate limits apply.
Related projects
Decision plugin for agent harnesses where Jev recommends tools, skills or owners and scores output, for DeepSeek Harness and iPolloWork.
Agent skillbetaTypeScriptPortable decision layer for agent harnesses: Jev picks from host-supplied capabilities while the host keeps permissions and execution.
ProjectbetaTypeScriptLocal Python MCP service that keeps the skill catalog out of the agent's context and lets Jev pick which SKILL.md files to load.
MCP serverbetaPythonCodex skill and Python CLI that recommends one installed skill per task using Jev Choice ranking and Noul fit checks, or returns none.
Agent skillbetaPythonLocal gateway, MCP server and Claude Code hooks that let coding agents delegate explicit choices, classification and scoring to Jev.
MCP serverbetaTypeScriptStandard-library MCP server giving Codex advisory Jev tools for role and resource selection, with local validation and 85 offline tests.
MCP serverbetaPython