Chrome for Developers

WebMCP is available for early preview → goo.gle/4rML2O9 WebMCP aims to provide a standard way for exposing structured tools, ensuring AI agents can perform actions on your side with increased speed, reliability, and precision.

🤖 Make your website agent-ready with WebMCP, the new protocol for reliable, structured AI → goo.gle/3ZWqkzu Join @andreban and Francois Beaufort on @WeAreDevs as they dive into how WebMCP moves the web beyond human-only UI to high-performance agent architecture.

Personal Intelligence is rolling out to more users for free across the Gemini app and Gemini in @GoogleChrome in the U.S. Access smarter responses uniquely relevant to you if you choose to connect your @Google apps like Search, @Gmail, @GooglePhotos, and @YouTube.🧵

@hansonjacke The ability to seamlessly hand off your active browser tabs to an AI assistant without losing any of your local state is going to completely change how we approach complex UI debugging.

Chrome 146 Autoconnect MCP - Game Changer! 🚀 Just discovered Chrome 146's built-in MCP autoconnect feature. This is HUGE for AI-agent automation! 🔥 What's New: • No need to launch separate Chrome instances • Auto-connect to your active browser session via MCP • Direct access to your existing tabs & workflows 🛠️ How to Enable: 1. Open chrome://inspect/#remote-debugging 2. Enable Remote debugging 3. Install chrome-devtools-mcp or agent-browser skill 4. Your AI agent can now: - Read page content - Click & interact - Inject JavaScript - Debug your active session Seamless switching between manual and AI-assisted debugging. The future is here! 🫡 Credits: @ChromiumDev @xpasky @wey_gu #Chrome146 #MCP #DevTools #AIAutomation #Chrome

@hansonjacke Exposing the browser's entire DOM and execution environment directly to an AI assistant through MCP is the ultimate frontend automation unlock.

Chrome MCP (Model Context Protocol) is amazing! 🚀 Just successfully posted this using Chrome DevTools MCP. It enables: 📸 Screenshot capture 🔍 Page structure analysis 🖱️ Click, input & user interactions ⚡ JavaScript execution 🌐 Browser automation It's like an AI cockpit for browsers! The MCP protocol connects AI assistants with browser capabilities. #ChromeMCP #DevTools #Automation

📐 Building corner-shape into the browser was a massive learning experience. Beyond just the code, this deep dive shares the technical hurdles and lessons discovered while implementing native curved corners → goo.gle/46DW9kz

@kode11 It really does feel like the missing piece of the puzzle, giving agents the real-time context they need to finally navigate the web exactly like a user would.

WebMCP is genuinely exciting. Agents that can inspect and interact with live browser state through a standardized protocol — this is the missing piece for agentic workflows that need to operate in the real web, not just API-land. The Model Context Tool Inspector is a smart move for debugging too.

Take a bigger slice of the agentic web this #PiDay by shipping a literal pizza pie → goo.gle/3PaAVEL To try: enable chrome://flags/#enable-webmcp-testing & install the Model Context Tool Inspector extension. Share what "pi" you’d build below! 🍕

@truechatdata It is incredibly easy to get distracted by the latest model capabilities, but grounding the project in the actual user outcome is exactly what separates a neat demo from a production-ready feature.

@ChromiumDev This framing is useful because a lot of teams jump straight to model choice and skip the earlier question of whether the prediction will actually change a decision. The lifecycle starts with the intervention, not the model.

🔄 Manage the machine learning lifecycle → goo.gle/4aCvTJL The latest Learn AI module breaks down the iterative cycle of Predictive AI, from defining your opportunity to final deployment. Learn how to determine if predictive systems are the right tool for your use case.

@andreban The transition from vibe to architecture is exactly where deep performance profiling and a ruthless testing strategy become your best friends.

Vibe coding is the ultimate disposable medium for thought. 🌊 But don't mistake "polish" for production. To ship, you must flip the switch from Vibe Checker to Lead Architect. 🏗️ My breakdown on moving from Vibe to Production: bandarra.me/posts/vibe-to-…

@leonliuzx The best kind. 😎

it's your vibe "pi"

Building a Chrome Extension doesn't mean you have to ditch your favorite framework. However, frameworks relying on SSR or external CDNs might hit a wall with Manifest V3's security policies. Learn how to pick the right stack for your extension → goo.gle/4uAtFlV

@steipete @openclaw Letting the agent dynamically choose between an isolated sandbox and the live 'user' profile based on the context of the prompt is exactly the kind of intelligent routing this ecosystem has been waiting for.

New @openclaw beta is up: it comes with the new live browser control that Google added in latest Chrome! enable via chrome://inspect#remote-debugging Your clanker will know when to use what, or you can ast it. new "user" profile session is there! developer.chrome.com/blog/chrome-de…

@Kyriakos_Pelek The possibilites are endless.

@ChromiumDev WebMCP ideas starting to look interesting

@LanYunfeng64 and the future of the web never looked better.

@ChromiumDev this isn't just about browser extensions. that's not tooling, that's the browser becoming an agentic operating system.

@TryArcade @nearestnabors @MHablich This is a great resource for anyone looking to up their browser's game ⚒️

Every web developer's agent workflow: copy console error, paste into agent, wait, repeat. Chrome DevTools MCP fixes this: ~25 tools that give your coding agent direct access to the browser. @nearestnabors talked to @MHablich about how they built it. youtu.be/CIOcjjgl9MI

@andreban It's ex-za-ctly what the worlds needs 🍕

Love the WebMCP zaMaker demo! It's cool to ask an agent "Pepperoni pizza for 3* and see the agent figure out the size, base, and toppings all by itself !

@addyosmani Giving an agent native access to cross-tab state and Workspace APIs fundamentally changes how we handle context in the browser.

Gemini in Chrome is rolling out to more people in more languages soon. It can work across tabs, schedule meetings in Calendar, send emails in GMail and understand YouTube videos.

Announcing big changes to Gemini in Chrome - agentic browsing with Auto-browse, Nano Banana & more! 🚀

@stevoperisic Markdown and text files are great when it's about the agent only reading the content. Tools exposed via MCP or WebMCP allows the agent to effectively action on behalf of the user.

@ChromiumDev Ok so simple md files or text files are not effective. But, making your API layer offer both standard https and mcp capabilities is the way to go?

The shift from legacy pages to 'agent-friendly' architecture is officially here 🌐

@inflectivAI @matthiasrohmer We’re excited to see how you use these agent-aware capabilities to tackle real-world problems!

@ChromiumDev @matthiasrohmer This collapses the barrier between coding and execution. Agents move from suggestions to active problem-solving in live environments.

🚨 NEWSFLASH🚨 The barrier between AI agents & runtime applications has officially collapsed. Sources confirm the new DevTools MCP server now allows models to debug, trace, & fix issues directly in Chrome. @matthiasrohmer has the full report → goo.gle/4qwBzKu