Jev in production › Model and agent routing

Agent Squad

Multi-agent orchestration framework, formerly awslabs/agent-squad, ships JevClassifier in TypeScript and Python, which sends agent descriptions to Jev as one choice question and routes each turn to the winning agent with its confidence.

Open on GitHub ↗

The source publishes no measured number.
Use
Model and agent routing
Industry
AI infrastructure
Form
Open-source tool
Stage
In production
Listed
2026-09-23
Found via
github
Repository
2FastLabs/agent-squad
Stars
7,790
Forks
743
Last push
2026-10-08
Language
Python
License
Apache-2.0
Agent Squad screenshot
img/flow.jpg in the 2FastLabs/agent-squad README, Apache-2.0; shown from GitHub.

The README opens with

Agent Squad Flexible, lightweight open-source framework for orchestrating multiple AI agents — in the cloud with Python and TypeScript, and now on device with Swift.

JevClassifier — typed decision routing in TypeScript & Python

Route with a decision model instead of an LLM: calibrated confidence, sub-second decisions, one-line drop-in — in both runtimes. See what's included ↓ · Jev docs →

On-device agent orchestration for iPhone, iPad, and Mac — agents, MCP tools, realtime voice, and tracing, running entirely on device. See what's included ↓ · Swift README →

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for model and agent routing