WebMCP is a proposed web standard that lets a website expose structured "tools" to AI agents running in the browser. Instead of an agent guessing which buttons to click, the site tells it exactly what actions are available, what inputs they need and what they do.
For businesses, WebMCP is an early look at how websites may work with AI agents that shop, book and fill in forms for people. It's still experimental, so most sites don't need it yet. But it's worth understanding now if your site depends on forms, search, booking or checkout.
What is WebMCP?
Chrome's WebMCP documentation describes it as a proposed web standard to help you build and expose structured tools for AI agents, using JavaScript and annotated HTML forms.
The WebMCP specification is a Draft Community Group Report published by the W3C's Web Machine Learning Community Group. Its editors come from Microsoft and Google. The spec itself states that it is not a W3C Standard and is not on the W3C Standards Track. In other words, it's a serious proposal, not a finished standard.
The name borrows from the Model Context Protocol (MCP), the open standard for connecting AI applications to tools and data. WebMCP shares MCP's idea of tools and input schemas, but the spec says it is designed to complement, not replace, backend protocols like MCP.
Why WebMCP matters
Agents are bad at guessing. Today, a browser agent usually reads the page and tries to act like a person would. That can fail on complex forms or custom widgets. Chrome's origin trial announcement gives examples like telling a single full-name field apart from separate first and last name fields.
Major browsers are testing it. Chrome opened a WebMCP origin trial in Chrome 149, announced in June 2026. The spec's implementation status page also lists an origin trial in Edge 150. It lists no status yet for Firefox or Safari.
It connects to agentic commerce. As AI agents start completing tasks for users, sites that make those tasks easy and reliable may have an advantage in agentic commerce. Google's Chrome at I/O 2026 post said brands including Expedia, Booking.com, Shopify and Target were already experimenting.
How WebMCP works
The imperative API
Developers register tools in JavaScript. In the current draft, the API lives on document.modelContext. A tool has a name, a description, an input schema and an execute function that runs when an agent calls it. Some articles and examples refer to navigator.modelContext instead, so check the latest spec before building.
The declarative API
Ordinary HTML forms can become tools by adding attributes. The declarative API explainer describes toolname and tooldescription attributes on the form. By default, the agent fills in the form and the user still clicks submit. A toolautosubmit attribute lets the agent submit on the user's behalf.
Scope and lifetime
WebMCP tools only exist while the page is open. Chrome's guide on when to use WebMCP and MCP says that once the user navigates away or closes the tab, the agent can't access the site or take actions. A regular MCP server, by contrast, runs persistently on a server.
Security hints
Chrome's tool security guidance recommends marking high-stakes tools, such as booking or payments, with consequentialHint so the agent or browser can ask the user to confirm. It also describes readOnlyHint and untrustedContentHint, and warns that it's impossible to guarantee safety inside an LLM.
How to prepare for WebMCP
- Start with clean HTML forms. Clear labels and logical field names help people, assistive tech and agents today, and they make declarative tools easier later.
- List your key tasks. Identify the actions customers most want to complete, such as search, quote requests or bookings.
- Experiment safely. Developers can test locally with the chrome://flags/#enable-webmcp-testing flag or join the origin trial. The API is subject to change.
- Require confirmation for big actions. Keep a person in the loop for payments, bookings and account changes.
- Keep your core content crawlable. WebMCP helps agents act on a page. It doesn't replace technical SEO or the content AI search uses to answer questions.
Common mistakes and misconceptions
"WebMCP is an official W3C standard." It isn't. The spec is a Community Group draft and says it is not on the W3C Standards Track.
"WebMCP is the same as MCP." They share concepts, but MCP connects agents to servers anywhere, while WebMCP tools live in a browser tab and disappear when it closes.
"WebMCP will boost my rankings." None of the official sources describe it as a search ranking factor. It's about agents completing tasks, not about being found.
"Agents can do anything once tools exist." By default, declarative form tools still leave the final submit to the user, and the security guidance encourages confirmation for consequential actions.
WebMCP vs. MCP
Chrome frames the two as partners, not opponents. Use MCP for core logic and background tasks that agents can reach anytime. Use WebMCP for in-browser interactions while a user is on your site. Many businesses may eventually use both.
How Butter approaches WebMCP
We don't build AI agents, and WebMCP implementation is developer work rather than part of our plans. Agent readiness does overlap with the work we already do. Our GEO plan covers AI prompt testing, indexing strategy and schema markup guidance, which help AI systems find and read your key facts. On the GEO+SEO plan, technical SEO fixes for broken links and crawl issues make pages easier for any automated visitor to use.
When we review a site, we flag pricing, service details or forms that an agent would struggle with, and include those in our on-page recommendations. That makes a good starting point if your developers later decide to add WebMCP tools. Our GEO service page has the details.
Frequently asked questions
Is WebMCP available now?
Only experimentally. Chrome opened an origin trial in Chrome 149, and the spec's status page lists an origin trial in Edge 150. The API is still a draft and may change.
Who is developing WebMCP?
The spec is published by the W3C's Web Machine Learning Community Group, with editors from Google and Microsoft. It is a Community Group draft, not a W3C Standard.
Do I need WebMCP for SEO?
No. WebMCP helps browser agents take actions on a page. It isn't described as a ranking factor, and crawlable, well-structured content still matters most for search and AI visibility.
What's the difference between WebMCP and MCP?
MCP connects AI agents to servers and data from anywhere, at any time. WebMCP exposes tools from a live web page to the browser's agent, and those tools disappear when the user leaves the page.