Jev in production › Model and agent routing
Generative UI engine routes layout decisions to Jev instead of streaming LLM tokens; author claims responses of 6.5-25ms versus seconds-long token streaming (author).
Stream-first interactive visual runtime for AI agents — 60fps morphdom DOM reconciliation, streaming HTML sanitizer, and WebGL/Canvas widget bridge.
Render live-streaming interactive calculators, 3D scenes, charts, and explorable explanations inside Claude Artifacts, ChatGPT Canvas, Gemini, and custom agent webviews with zero flicker.
Quickstart · What it makes · Install · How it works · Commands · Limits
Naively setting innerHTML on an LLM token stream causes severe layout thrashing, destroys active / state on every chunk, and opens XSS vectors. design-os-generative-ui provides a stream-healer, sanitizer, and 60fps morphdom reconciler (createStreamRenderer + DesignOSBridge) purpose-built for AI agent artifacts.
For the project's own README, linking back here: