Does Jev generate the chat reply or only pick the model?

Jev only picks the option whose description best fits, returning choice, confidence and probabilities. The actual answer is generated by the mapped Databricks FMAPI endpoint using the app service principal.

Can I define my own routing options and target models?

Yes. You define labels like Coding or General with plain-English descriptions and target models. The UI manages options and /router/chat returns which model answered plus routing and inference timings.

What credentials does local run versus Databricks deploy need?

Local run needs JEV_API_KEY plus Databricks host and token for FMAPI access, then uvicorn app:app on port 8080. Deploying needs the key in a secret, app.yaml wiring, and CAN QUERY on endpoints.

What data leaves my workspace when a message is routed?

The conversation transcript used for the routing decision is sent to the external JEV endpoint. The digest warns to review data-handling requirements before routing sensitive content.

What license does the Jev Smart Router use, and what stack is it built on?

MIT. It is a FastAPI app packaged for Databricks Apps, with a single-page UI, and it calls Jev's native choice question type then invokes a Databricks FMAPI serving endpoint over the raw REST invocations route.

How are reasoning models handled when the routed endpoint returns no content?

Endpoints that put text in reasoning_content instead of content are handled with generous max_tokens and a reasoning_content fallback. Inference uses the raw REST /serving-endpoints/{model}/invocations call rather than the SDK's typed query() method, which has dict-handling quirks.

Which environment variables tune the Jev routing decision step?

JEV_API_KEY is required; JEV_API_URL defaults to the System One endpoint, JEV_MODEL to jev-latest and JEV_TIMEOUT to 30 seconds, while JEV_LOG_PROMPT=1 logs routing decisions to stdout. Set the key from a secret, never hardcoded.

CommunityJev Agents & MCP

Jev Smart Router

Databricks App that sends each chat message to TypeSafe Jev choice routing, then runs inference on the winning Databricks FMAPI endpoint with confidence and per-option probabilities shown.

See highlights

Open github.com

Stars: 5
Language: Python
License: MIT
Published by: Community
Author: rmosleydb
Last verified: Oct 2026

rmosleydb/jev-smart-router

User repository on GitHub

View on GitHub

JEV Smart Router — a Databricks App that uses TypeSafe JEV to pick which model answers each message, then runs inference on the chosen Databricks Foundation Model API endpoint.

Forks: 2
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

A Databricks App that uses TypeSafe JEV (System One) to decide which model should answer each message, then runs the real inference on the chosen Databricks Foundation Model API (FMAPI) endpoint.

Quoted from the project's README. Read the full README.

Tagscommunity awesomejev catalog-import

Works forJev for Classification Jev for Routing

Used inJev for Software & DevTools

Languages: Python

Highlights

Uses Jev choice type over option descriptions with confidence and probability breakdown.
Maps winning option to a databricks-* FMAPI endpoint invoked via service principal.
FastAPI endpoints for route-only decisions, routed chat, live models and health probe.
UI shows routed model, confidence tooltip, routing latency and inference time.

Watch out

Requires a TypeSafe key plus Databricks workspace access, and routing transcripts are sent to api.typesafe.ai. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

muse-jev-playbook

Agent-oriented playbook for using Jev as a cheap, fast decision layer that triages, classifies, scores, and gates work before expensive browser, research, retry, or subagent steps.

Jev Agents & MCP#community#awesomejev#catalog-import

Community

Bodila51/muse-jev-playbookPython

claude-jev

Coding-agent plugin that routes prompt classification, subagent tiering, edit rule checks, and fast Jev-backed compaction through small typed judgments.

Jev Agents & MCP#community#jevbest#catalog-import

Community

0x7067/claude-jevTypeScript

jevymarket

Polymarket bot that pairs a web-search researcher brief with Jev pricing, trading only where Jev probability clears the order book by a set edge.

Jev Agents & MCP#community#awesomejev#catalog-import

Community

markusbug/jevymarketPython

typesafe-skill-router

Opt-in Hermes plugin that routes each turn to one skill: a wide Jev Choice plus request checks gate a narrow fits round, injecting a single ignorable hint line.

Jev Agents & MCP#community#jevbest#catalog-import

Community

DECRUX9812/typesafe-skill-routerPython

tink-route

CLI that routes a task to one specialist skill from a Tink library using TypeSafe Jev, verifies the winner with tink mount, then prints it inline or as a mounted path.

Jev Agents & MCP#community#awesomejev#catalog-import

Community

jon-devlapaz/tink-routePython

hermes-jev-plugin

Hermes Agent plugin exposing Jev as four coding-workflow tools for parallel judgments, binary checks, option routing and rubric scoring, with a bundled question-design skill.

Jev Agents & MCP#community#jevbest#catalog-import

Community

ajensenwaud/hermes-jev-pluginPython

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai