The State of WebMCP
Not what WebMCP is, that's covered elsewhere. This is where things actually stand right now: what's shipped, what's still an open question, and what to do about it today. We'll keep this updated as the picture changes.
TL;DR
- • Native
document.modelContextis running as a public origin trial in Chrome and Edge, versions 149 through 156, started May 2026. - • Every other browser (Firefox, Safari, and Chrome/Edge outside the trial window) needs the community polyfill, which most WebMCP tooling loads automatically.
- • Lighthouse added a real
agentic-browsingscoring category, so "is my site agent-ready" now has a measurable, repeatable answer. - • Whether the origin trial converts to a permanent, unflagged API, whether other browser vendors commit to it, and whether it ever factors into search ranking are all still open questions. Nobody, including Google, has answered them yet.
Where the spec stands
Google shipped document.modelContextas a public origin trial starting in Chrome 149 (May 2026), running through Chrome 156. Edge follows Chromium's release schedule and runs the same trial. That's the only place the native API exists today, everywhere else the community WebMCP polyfill fills the gap by reimplementing the same interface in JavaScript.
An origin trial is explicitly experimental: Chrome can extend it, ship it as a standard permanent feature, change the API shape, or drop it, and has done all four with other origin trials in the past. Building on the polyfill today means your tools work everywhere regardless of which of those happens.
The ecosystem so far
The polyfill
The mcp-b/webmcppolyfill is the de facto compatibility layer right now, it's what makes WebMCP usable on non-Chrome browsers and pre-149 Chrome today instead of waiting on the trial.
Lighthouse scoring
Lighthouse 13.4's agentic-browsing category gives site owners a concrete, automatable way to check WebMCP compliance, six equally-weighted audits covering tool registration, schema validity, form coverage, accessibility tree structure, layout stability, and llms.txt.
Managed tooling
Platforms like Aigentably exist because hand-writing registerTool() calls, schema validation, origin scoping, and cross-browser fallbacks for every page of a site is real work. Expect more tooling here as adoption grows, not less.
What's still genuinely open
Being direct about what we don't know matters more here than pretending certainty:
- Does the origin trial graduate?Origin trials are trials. Chrome hasn't committed to shipping WebMCP as a permanent, unflagged API past Chrome 156.
- Do other browser vendors adopt it? Firefox and Safari have made no public commitment to implement
document.modelContextnatively. The polyfill is doing that work in the meantime, indefinitely if native support never arrives elsewhere. - Does it become a ranking signal?Core Web Vitals followed the pattern of "Google measures it, then it starts to matter for visibility." Lighthouse now measures agentic browsing. Whether that pattern repeats is a bet, not a fact.
- Which AI agents actually query it in production? An agent has to check
document.modelContextfor it to matter at all. Agent vendors' own adoption timelines are largely undisclosed.
What to do about it now
You don't need to wait for these questions to resolve. The polyfill means tools you register today work in every major browser right now, native support in Chrome/Edge is a bonus, not a prerequisite. If the origin trial doesn't graduate or other vendors never adopt it, you've lost nothing, the polyfill keeps working. If it does become the standard, you're already there.
The practical first step is small: pick two or three actions on your site an agent would plausibly want to take, e.g. searching, adding to cart, checking status, and expose them as WebMCP tools. See what happens.
Related reading
What is WebMCP?
The definitional explainer, start here if you're new to it
Browser support breakdown
Chrome/Edge origin trial, polyfill coverage, version by version
WebMCP vs MCP
Two different protocols, two different layers, cleared up
WebMCP vs browser automation
Why scraping and automation are the wrong approach for agents
Lighthouse agentic browsing score
How Chrome's Lighthouse now audits WebMCP compliance
schema.org markup as a WebMCP shortcut
Most sites already have the JSON-LD a read-only tool needs
Security: tool poisoning & prompt injection
The attack surface WebMCP tools introduce, and how to close it
Implementation examples
Copy-paste registerTool() patterns for common use cases
WebMCP for e-commerce
Making Shopify and retail sites agent-ready
Make your site agent-ready today
No need to wait for the open questions to resolve. Aigentably handles the WebMCP infrastructure, native and polyfilled, so your tools work regardless of how this shakes out.
Get started free