A2UI: A Protocol for Agent-Driven Interfaces
4 days ago
- #AI Agents
- #Open Source
- #User Interfaces
- A2UI enables AI agents to generate rich, interactive user interfaces that render natively across web, mobile, and desktop without executing arbitrary code.
- Status: Early Stage Public Preview (v0.8), Apache 2.0 licensed, created by Google with contributions from CopilotKit and the open source community.
- Solves the problem of how AI agents can safely send rich UIs across trust boundaries using declarative component descriptions.
- Features: Secure by Design (pre-approved components), LLM-Friendly (flat, streaming JSON), Framework-Agnostic, Progressive Rendering.
- Get Started: Run the restaurant finder demo, understand surfaces and components, integrate A2UI renderers, or dive into the technical specification.
- How It Works: User sends a message to an AI agent, which generates A2UI messages describing the UI, streams to the client, which renders using native components.
- Demos: Landscape Architect Demo (custom form generation), Custom Components (Interactive Charts & Maps), A2UI Composer (Widget Builder by CopilotKit).