I was adding generative UI and user input handling to my AI agents and kept repeating the same logic. So I turned it into AgenticGenUI — and now I’m open-sourcing it.

It’s a React library that lets you add generative UIs and capture interactions with just one line of code. It renders components from structured data and emits structured responses (tool calls) back to your agent.

Live demo: https://v0-open-source-library-creation.vercel.app GitHub: https://github.com/vivek100/AgenticGenUI

45+ components: MetricCard, MultiStepForm, KanbanBoard, ConfirmationCard, etc. Works with any LLM/runtime, CopilotKit-ready, fully typed, themed Would love feedback and contributors!