Jev in production › Model and agent routing

langchain-skill-router

Jev picks which skill handler takes each turn in a LangChain deepagents router; author reports 90% vs 88% accuracy at 4x fewer tokens per turn. (deyna256)

Open on GitHub ↗

90% vs 88%measured against a baseline, as published by the source
Use
Model and agent routing
Industry
Developer tools
Form
Open-source tool
Stage
In production
Plugs into
LangChain deepagents
Listed
2026-10-05
Found via
github
Repository
deyna256/langchain-skill-router
Stars
6
Forks
5
Last push
2026-10-05
Language
Python
License
MIT

The README opens with

A drop-in replacement for the deepagents SkillsMiddleware : on each user turn a fast judge decides which SKILL.md skills are needed, and only those are loaded, so a catalog of hundreds stays out of the prompt. Bring any judge: a hosted model, a self-hosted one, or plain rules. An adapter for Jev is included.

113.0k → 25.8k input tokens per turn on a 236-skill catalog, and the agent answered 90% of questions correctly against 88% with the whole catalog in the prompt. See the benchmark →

Badge

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

Listed in Jev in production

Also used for model and agent routing