orchestkit
Large Claude Code plugin of skills, agents and hooks that can optionally ask Jev to classify a session's work category.
Overview
OrchestKit is a large plugin for Claude Code and other hosts, with 110 skills, 36 agents and 162 hooks; Cursor, Codex, pi, Devin and Antigravity get subsets, documented in a measured host-support matrix. It covers a broad range of development workflows.
The Jev integration is optional: setting ORK_SESSION_CATEGORY_PROVIDER to jev (or shadow) with ORK_TYPESAFE_API_KEY makes the session-identity hook, which already asks a local Haiku process for a title and category, also ask TypeSafe's Jev (pinned to jev-1.13.0) one typed Choice over the same eight categories and criteria. Per the list summary, the result sets session colours when confidence meets a threshold. Shadow mode allows comparison without acting.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Which of eight work categories does this coding session belong to? | — |
How it works
- Session-identity hook runs at session start.
- Haiku proposes a title and category.
- With Jev enabled, ask one Choice over the eight categories.
- Apply the category and colour when confidence meets the threshold.
Strengths
- Optional second classifier with shadow mode.
- Pinned Jev version for stable results.
Caveats
- Jev is a minor optional feature in a large plugin.
- Requires a separate TypeSafe key variable.
Getting started
In Claude Code: /plugin marketplace add yonatangross/orchestkit, then set ORK_SESSION_CATEGORY_PROVIDER=jev and ORK_TYPESAFE_API_KEY.At a glance
- Type
- Project
- Status
- production
- Language
- TypeScript
- Runtime
- Claude Code plugin
- Has code
- yes
- Has tests
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
OrchestKit can optionally use Jev to classify coding sessions and set their colors when confidence meets a threshold.
logicrw/awesome-jev-projects
Frequently asked questions
What does Jev classify?
The work category of a coding session, from eight fixed categories.
Is Jev required?
No, it is optional and can run in shadow mode.
Which Jev version is used?
It is pinned to jev-1.13.0.
Related projects
Local-first macOS coding workspace that runs your agents and lets a local Laya or hosted Jev selector pick a route the host validates.
ServicebetaRustLocal-first, outcome-aware model router for coding harnesses that picks a capability tier per task with Jev and learns from history.
ProjectprototypeRustLocal guard layer for coding agents that blocks dangerous tool calls by deterministic rules, with optional Jev judgments.
CLIbetaTypeScriptLocal-first meeting transcription and conversation memory app whose live voice path runs its evaluations through Jev.
AppbetaRustIDE and browser extensions to find and install agent Skills and Subagents, with an opt-in Jev check when local detection can't decide.
Browser extensionproductionTypeScriptLocal desktop pal that verifies what a coding agent really did, with optional Laya or Jev double-checks of unclear test results.
ProjectbetaTypeScript