What does it do?
A website normally gives visitors pages, links and forms. WebMCP lets its developer also offer a small set of named actions, with a description of each action and the information it needs. A compatible browser assistant can then help a visitor use those actions within the open page.
Our first prototype is a contact-page action called “prepare an enquiry”. It can fill the visible name, email and message fields. The visitor still reads the result, acknowledges the privacy notice, completes the spam check and chooses whether to send it.
Why might that matter for your site?
People often arrive with a job in mind: find the right service, narrow a list or ask a question. A well-defined action can help them get there with fewer steps. The website’s own rules still decide what information is required and what happens next.
That clarity matters to the site owner too. The assistant gets a description of a specific task instead of guessing its way around buttons and fields. You can decide which actions to offer, keep sensitive steps with the visitor and leave the ordinary website available to everyone.
WebMCP is still experimental
WebMCP is an emerging browser API, not a feature every visitor can use yet. Browser support, permissions and the exact interface are still developing. Adding it does not make a site rank higher in search, replace accessibility work or remove the need for a good contact form.
The contact-form helper is deployed on one client site. In a local test, it filled the form without sending. We have not yet confirmed that a compatible browser finds and runs it. The normal form remains available for browsers without WebMCP.
The practical question
Which job should your site make easier?
A useful first step is to pick one task visitors already try to complete, then make that task clear and reliable. If you have one in mind, tell us what visitors are trying to do.
Tell us about your site