Ask HN: What Is the Point of WebMCP?
So I have a question that should be simple, but apparently is not. I have Chrome Canary with WebMCP enabled. I have the Model Context Tool extension installed. I have a WebMCP-enabled test app running in it.
The only thing the extension offers is connecting to the Gemini API via an API token and using that as the driver for the WebMCP app. I don't understand this use case. If I wanted my app to consume API tokens I would have directly coded interoperation with the major providers with one of the thousands libraries that di exactly that.
What I do have are subscriptions to OpenAI, Gemini, Claude, and GitHub Copilot. All the clients I use can connect to MCP servers. I thought: I have WebMCP enabled in Canary, so my browser will actually be an MCP server that gives access to the underlying WebMCP apps it is running. It turns out that no, of course that would be logical, but you can't do that. Chrome exposes an MCP server, but it is only for controlling Chrome itself, like Playwright. It does not provide any bridging to WebMCP.
I then thought there would be an extension that does this, but the Chrome extension ecosystem is full of bullshit extensions from fraudsters who vibe-coded their way into trying to capitalize on the AI craze. There are no extensions for that. All of them want you to subscribe, or give them API tokens. Many of them have non-working websites. One option I see mentioned online is running Chrome in debug mode and exposing a remote debugging port. This is exactly what WebMCP was supposedly invented to avoid, so running your browser in a special way to expose it to complete control cannot be the right way to do this. I'm perfectly able to already do that without WebMCP.
Am I understanding it wrong? I understand the proposal is in its infancy, but I expect the goal of it to be at least clear. I don't understand where it helps developers or users in any single way that wasn't already available before. The value of WebMCP becomes clearer when you think about non-developers trying to connect AI to their existing tools. Right now if a small business owner wants an AI agent to interact with their CRM or booking system, they need a developer to build custom integrations. A standardized protocol for web-based tool access could make that accessible to people without engineering teams. Whether WebMCP specifically is the right approach is debatable, but the problem it's trying to solve is real. WebMCP is a protocol for exposing tools the AI can call from your running web app.
the point isn't "consume API tokens", it's "let the AI do stuff in your app" (click buttons, fill forms, read DOM state). The Gemini integration is just the orchestrator for the example implementation. not the protocol Yes, but in practice, how do you connect AI to WebMCP? What is the point of a protocol that speaks MCP inside the browser if it's not reachable outside the browser? Many browsers now have AI agents built into them. For example, Chrome has Gemini-in-Chrome, and Atlas has ChatGPT. When a user is on a website and then opens the built-in browser AI agent, the AI agent will see the WebMCP tool calls for that website and be able to call them.