⭐️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 👀 github.com/ag-ui-protocol/ag… How it works & demo 👇

@LangChain Interrupt26 in NYC was amazing! 🗽 The energy of a sold out venue of agent builders from enterprises all over the world was so inspiring. @craigirwin and I spent the day demoing @CopilotKit and answering questions. The favorite topic: how your app can have automatic learning built in 🤯

That's a wrap on @LangChain Interrupt26 NYC - incredible keynote by @hwchase17! @nathan_tarbert & @craigirwin ran demos all day - showed how you can have everything you need for an agentic app using CopilotKit, including automatic learning - getting better with user interactions. Missed us? DM us for a demo 👇 go.copilotkit.ai/book-a-demo

Marc Benioff said "It's not the end of software." He's right, it's the end of software that makes humans do all the work. We've been saying this since April. For decades, we went to software. Now, software comes to us. → Agents do the work → Interfaces generate on the fly per task → Humans stay in the loop to steer, approve and decide This is what we're enabling anyone to do with @CopilotKit - the frontend stack for agents, on ANY app. The "interface revolution" won't be locked into any one vendor. It will run on an open stack. AG-UI is the protocol that makes this possible. I wrote about the two forces reshaping software right now: the headless web and agent-generated interfaces. Watch the talk, then read the article 👇

Excited to try this with kody.codes

OpenMuse is powered by TanStack AI 🌴

Two days... 2,000 GitHub stars! We're cooking some really cool stuff for OpenMuse. (Hint: Jev, Render, Learning...) Try it out 👇

1M impressions for OpenMuse 🎉

Had a great time demoing @CopilotKit at @AITinkerers demo day 🛠️ A highlight was showing OpenMuse live and finding out that people in the room were already using it. Want a demo? DM me. Thanks to @rjnmrry , @posthog, @veris_ai, @tavilyai 🙌

Who's in New York? I'm going to be giving a @CopilotKit demo at the @AITinkerers event. I'll even throw down some OpenMuse. Come down, hang out, and say hi. I'd love to find out what you're building! nyc.aitinkerers.org/p/septem…

New in OpenMuse: Automatic Learning The more your OpenMuse is used by users, the better it gets. Product trajectories from user-interactions are turned into skills for your agents. Try it: github.com/CopilotKit/openmu…

STOP making your agent learn only from the chat This automatic learning works across your entire product 👇🏻 Most AI strategies are optimizing the wrong data source. I've spent the last few months investigating CTOs, Chief AI Officers, and heads of product at companies deploying agents at real scale. There's a pattern I keep noticing: Everyone is investing heavily in learning from agent conversations. Meanwhile, 90% of the actual signal is sitting right next to the chat window, completely untouched. Your users don't just talk to your agent. They *use* your product. They click, hesitate, open pages, abandon flows, message a teammate, convert and churn. That behavioral surface is 10–100x richer than the chat conversation itself! And it's the context your agent needs to actually be useful in an enterprise setting. 👉 I’ve noticed that the team at @CopilotKit is one of the few actually building the infrastructure for this, called Intelligence. Here's how it works: > Analyzes product interactions and agent traces > Flags patterns and proposes new skills > Your agent ingests every skill you approve Now every user-interaction in your product becomes a learning signal. If you want to pull ahead right, you need to make your agents learn from the whole product, not just the chat window. Worth a look if you're thinking about where your AI stack needs to go in the next 12 months. Get early access to it here: go.copilotkit.ai/get-learnin…

1,200 GitHub stars for 𝙾𝚙𝚎𝚗𝙼𝚞𝚜𝚎! 🚀 There's been an incredible reaction in less than 24 hours. > Copy it > Customize it > Make it your own Doing my best to answer all questions. Try 𝙾𝚙𝚎𝚗𝙼𝚞𝚜𝚎 now 👇

🎉 Introducing 𝙾𝚙𝚎𝚗𝙼𝚞𝚜𝚎 An open source, self-hostable personal assistant that works with any agent harness. Includes: - Computer use: browser, terminal & files - Connectors for your personal apps - Ideas, goals & progress tracking - Built for Mobile and Web Repo → github.com/CopilotKit/OpenMu… Powered by @CopilotKit and AG-UI. Clone this template and customize it however you want.

Introducing Open Muse An open source personal assistant that works with ANY agent harness. Available on Mobile and Web! Clone it. Build on top of it. 👾 Repo: github.com/CopilotKit/OpenMu…

between @idosal1’s big-brother-like appearance & @TobinSouth’s impromptu panel, MCP Apps Night was so much fun😂 big thanks to our speakers, community demoers, & sponsors! can’t wait for the next one💥💥 @tyler_cpk @khandrewone @opencolin @JimTheo_dev @weebhek

The agent↔user boundary produces more than text. It produces UI the agent rendered, approvals you gave, and files it wrote. A thread should be restored to it's original state, even on a different device.

Rich Threads in CopilotKit Intelligence capture the full interaction surface. Not just messages. Application state over time, persisted and resumable across sessions: - Generative UI, not just text responses - Approvals, edits and guided steps - Agent and UI stay synchronized - Uploads, artifacts and outputs preserved Durable streams and multi-device support recover from disconnects, even on mobile networks. Book a demo → go.copilotkit.ai/book-a-demo