User repository on GitHub
Adaptive model and reasoning router for Codex agents, with optional Jev-powered task assessment.
- Forks
- 0
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Agents & MCP
Codex Route Advisor
A skill for Codex agents that splits development work into Bounded Tasks and produces a Dispatch Plan recommending a model, reasoning effort and tools per task, with optional Jev-powered assessment and an Agent fallback when Jev is unavailable or low-confidence. MIT-licensed.
- Stars
- 6
- Language
- JavaScript
- License
- MIT
- Reading time
- 2 min
- Published by
- Community
- Author
- kenming
- Last verified
- Oct 2026
Used inJev for Software & DevTools
- Languages
- JavaScript
- Capabilities
- Open source
Highlights
- Splits only where needed: bounded tasks are formed so a local implementation and its focused test usually stay together.
- Allocates a model, reasoning effort and tools per task without escalating every task because one is difficult.
- Makes execution relationships explicit: dependencies, parallel groups and conditional work.
- Optional Jev assessment over an independent backend, with Agent fallback when Jev is unavailable or low-confidence.
- Advisor owns planning; the active Coordinator owns execution, delegation, retries and final validation.
Watch out
README reviewed 2026-10-17; the digest truncates the install section mid-command, so the exact Skills CLI invocation and the Jev backend configuration are not captured. The README declares no minimum Node version — 24.x is only the development baseline — and reports no routing benchmarks.
Questions about this entry
Short answers drawn from the linked project or page.
A skill for Codex agents that breaks development work into bounded tasks and returns an actionable Dispatch Plan: which model, how much reasoning effort and which tools each task should get, plus the dependencies between tasks.
As an optional independent backend for task assessment. The README says an Agent fallback takes over when Jev is unavailable or low-confidence, so the skill still produces a plan without it; the backend configuration details are not in the digest excerpt.
Through the Skills CLI, which requires Node.js and npm, run from your project directory, with Node 24.x as the development baseline. The digest truncates the install block, so the full command should be taken from the README.
Not necessarily: Jev and an API key are described as optional, because the Agent fallback can assess tasks when Jev is not available. What the fallback does differently is not documented in the captured text.
It does not switch the main chat model or control a Provider, Model Picker or Proxy. Planning belongs to the Advisor; the active Coordinator executes and validates, delegating and retrying as needed, per the README's separation of responsibilities.
The README's rationale: a local implementation and its focused test usually stay together, a difficult task does not force everything above it into a higher tier, and explicit parallel and conditional groups make execution relationships visible in the plan.
Kenming built it. Metadata: JavaScript, MIT, 6 stars, 0 forks, created and pushed 2026-10-08 — a single-day repository at capture time. The README is bilingual with a Traditional Chinese edition.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
Facebook: Intake Router keeps diagnosis out of the answer space at the front desk.
FacebookClinicnotes
X: OpenRouter Decision Model Rankings
X: herdr + jev, an agent router
X: A group chat bot with decision room routing
More Jev resources like this
Launchers for Claude Code, Codex, Grok CLI, and opencode that ask Jev which model-and-effort tier each human turn needs, rewriting only the model field while the real CLI runs unchanged.
Jev Agents & MCP#community#javascript#routing
Community
flaviusapop/jev-routerJavaScript
flaviusapop/jev-routerJavaScriptA local MCP server that lets Codex delegate routine drafting, analysis and code proposals to Gemini while TypeSafe Jev reviews each draft against explicit criteria; Gemini runs on included Google AI Pro quota and Jev needs its own TypeSafe account.
Jev Agents & MCP#community#github#mcp
A Claude Code plugin from alexei-led that picks the model and effort tier, micro to high, for each turn, advised by Jev, Cloudflare's Clef family, OpenAI, or a local Ollama classifier. It changes only model and effort, and it never proxies Anthropic traffic.
Jev Tools & CLIs#community#github#claude-code
Community
alexei-led/claude-routerJavaScript
alexei-led/claude-routerJavaScriptA Claude Code mod that draws a live cockpit above the prompt in CLI and Desktop: plan bars, subagent strips, usage limits with reset countdowns, and the model route jev-model-router picked with its confidence. It watches and draws only.
Jev Tools & CLIs#community#github#javascript
Community
nvr0x5/claude-deckJavaScript
nvr0x5/claude-deckJavaScriptClaude Code and Codex skills plus a study course for building with Jev, TypeSafe's System One model. The materials pull from Jev's canonical cookbooks and docs, and the Ratings folder lists rated Jev projects and an evaluation methodology.
Jev Agents & MCP#community#github#skills
Community
tiffygk/jev-modeHTML
tiffygk/jev-modeHTMLA model router for the Codex CLI: on every user turn Jev judges how hard the task is and a YAML policy picks one of three lanes — flash, craft or forge — so premium quota goes to the prompts that need it.
Jev Agents & MCP#community#typescript#codex
Community
romanmeclazcke/codex-siftTypeScript
romanmeclazcke/codex-siftTypeScriptFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.