About HiA2UI

HiA2UI is an independent, community-maintained documentation and learning project for the Agent-to-User Interface (A2UI) protocol. It is not the official A2UI website and is not operated by Google or the official A2UI maintainers.

For authoritative specifications, releases, and governance information, visit the official A2UI repository.

Our Mission

To give AI agents a rich, native, and safe way to communicate with humans.

For too long, AI has been trapped in the “chatbox”—a narrow window of text bubbles. We believe that for AI to be truly helpful, it needs to be able to show, not just tell. It needs to render charts, booking forms, maps, and interactive tools directly in the conversation.

But it must do so safely. We reject the idea of AI generating raw code (HTML/JS) on the fly, which is a security nightmare. Instead, we champion A2UI: a declarative, data-driven protocol that keeps humans in control.

Read our Manifesto: Why We Built A2UI

The “Handshake” Philosophy

We view A2UI as a diplomatic handshake between two very different worlds:

** The Probabilistic World of AI**: Creative, messy, unpredictable.
** The Deterministic World of UI**: Strict, structured, pixel-perfect.

Our protocol acts as the translator, allowing these two worlds to collaborate without breaking each other.

Core Values

🔓 Open & Neutral

Protocols should be free. A2UI is Apache 2.0 licensed and will never be locked behind a paywall. We support all LLMs (OpenAI, Anthropic, Gemini, Llama) and all frameoworks (React, Vue, Flutter).

🛡️ Security First

We believe "Eval is Evil." Our entire architecture is designed to prevent prompt injection and XSS attacks by strictly separating data from code.

⚡ Native Experience

Webviews and iFrames are cheats. We strive for 60FPS, fully native experiences that feel indistinguishable from handwritten applications.

🤝 Developer Centric

We build tools we want to use. Typed SDKs, great documentation, and a focus on Developer Experience (DX) above all else.

Contact Us

Whether you want to contribute to the protocol, report a bug, or just say hello, we’d love to hear from you.

About Us