⭐️Introducing AG-UI;
The Agent-User Interaction protocol 👾
Bring your AI agents into Frontend applications, and let them interact with users.
Launching with day-0 integrations with LangGraph, CrewAI, Mastra, and AG2 - with more partnerships on the way 👀
https://t.co/CQ5bvcJrSB
How it works & demo 👇
#MicrosoftBuild day 2 highlights:
✤ meeting @swyx & getting excited for @aiDotEngineer world’s fair in july
✻ seeing the @CopilotKit x MSFT Agent Harness on stage
✸ the beautiful weather at fort mason today!
Generative UI has just gone beyond traditional web. ✨
We got it working in Slack, Teams, Google Chat, Discord, and beyond.
Going live tomorrow to walk through our challenges and how we got around them.
If this is something you need, we'd love to hear from you.
We are opening early access today.
Early access form 👉 https://t.co/a91EdbZ1L7
Register for the event 👉 https://t.co/BIAot2VbcR
Chat was the first UI for AI. Generative UI is the next one.
CopilotKit lets your agent render real React components,
Call frontend tools, and drive the actual app.
Open source, built on AG-UI.
Start building here 👇
https://t.co/mFTyD9jvjp
💫 CopilotKit has quietly become a powerhouse of an SDK over the last few months.
@CopilotKit provides the building blocks for developers to build rich fullstack agent applications & Generative UI.
Including, but not limited to:
• Chat UI (pre-built or headless)
• Generative UI
• Shared State (between agent & frontend)
• Frontend Tool Calls
• Human-in-the-Loop
• Streaming & Reasoning UI
and more!
We created the AG-UI protocol, which lets you connect to any agentic backend like... 👇
LangGraph, ADK, Mastra, Strands, MAF, PydanticAI, LlamaIndex, etc.
The core promise of CopilotKit (the building blocks for fullstack agent apps) still fills a big gap in the ecosystem that many have yet to catch up on.
But over the last few months we've quietly upgraded beyond where most of the space sits today — and we now have the richest ecosystem out there.
I explain why below. Here's the TL;DR
1. Industry leaders in Generative UI
2. Frontends SDKs for React, React Native, Angular, Slack, Google Chat, Microsoft Teams
3. Solutions for Threads, Persistence & Self-Learning
We're still here to serve the 90% of the ecosystem that needs CopilotKit's core building blocks — but we're well ahead of the curve for where things are headed.
Soon, all UI will be AI!
🎉 Introducing CopilotKit (v1.59)
All the building blocks for fullstack agent applications & generative UI
Built on top of the AG-UI protocol — so it works with LangGraph, Mastra, ADK, Strands, MAF, PydanticAI, Agno, LlamaIndex, CrewAI, AG2 & Claude Agents.
✨ Includes:
• Shared State (between agent & frontend)
• Chat UI (pre-built & headless)
• Frontend Tool Calls
• Generative UI
• Human-in-the-Loop
• MCP Apps
CopilotKit is the most comprehensive toolkit for building Cursor- and Claude-like experiences inside your own product.
v1.59.2 introduces:
- Support for threads
- CopilotKit Vue packages
- React Native SDK
- v2 rendering hooks
And more!
All open source (32k stars 🌟)
Thanks to @copilotkit for being a Gold Sponsor at CascadiaJS 2026 this year ❤️
CopilotKit is the agentic frontend stack & the company behind the AG-UI protocol. This protocol has been adopted by Google, Microsoft, AWS, LangChain & more, making it the easiest path for bringing AI functionality (like Generative UI) to full stack applications.
Learn more at: https://t.co/cbv6aDPEZe
You’ve built an agent, but have you built an agent that communicates with other agents over the internet? What about an agent that generates a custom UI on-demand?
I'm super excited to announce that @a2anetcom are co-hosting London's first A2A protocol and generative UI hackathon with @Google and @CopilotKit on Saturday 13th June!
The A2A track is not your typical hackathon track. Your task will be to build three specific agents involved in customer service: a personal agent, a customer service agent, and a research agent.
This track will be judged on a held-out test set. Half of your score will come from how well your agents work together, the other half will come from how well each of your agents works with two other agents. Therefore, you need to build agents that work well with each other, and other agents!
The generative UI track is an exciting, open-ended hackathon track with infinite possibilities. For this your task is to build a product that features at least one agent and generative UI.
This track will be judged on a number of different criteria, like originality, economic value, technical difficulty, and use of generative UI!
This is NOT one you’re going to want to miss! Check the comments for the Luma page 👇✨
AG-UI is one of the more practical protocols that came out recently.
While MCP gives agents tools and A2A lets agents talk to each other, AG-UI focuses on the missing piece: letting agents properly interact with the user interface in real time.
It creates a clean two-way connection between your backend agent and the frontend, so the agent can render components, update state, and handle human-in-the-loop flows without hacky workarounds.
We’ve been using it in production and it makes building agentic UIs much cleaner.
@CopilotKit
#MicrosoftBuild day 1 highlights:
seeing our @CopilotKit x MSFT Agent Harness demo come to life at the Agents & Apps booth
chatting with @jacobandreou & @OmarShahine after meeting them at Builders on Tap last night🍻
come see the demo yourself at the Festival Pavillion!
CopilotKit ships the primitives on top of it:
- useAgent for shared state
- useFrontendTool for agent-rendered components
- Headless UI when you want full design control