WebMCP is available for early preview • Hacker News Board

Re: WebMCP is available for early preview

#171

Post by wongarsu » Mon, Mar 02, 2026, 10:25 AM UTC

hmdai wrote: post #161

Genuine question, why can't this be done via an API that the agents call? there are already established ways to call APIs on behalf of the user. Seems to me that the agent is loading a web app just to be able to access it's apis, what am i missng?

Yeah, we could have just standarized around a path to api specs. Maybe .well-known/openapi.yaml

Maybe it's cynical, but the best reason I can come up with is that 'established common url for api specs' does not sound nearly as cool on a CV or when talking about the next promotion as 'invented WebMCP'. And for those implementing it on their websites 'we implemented WebMCP' is again much more 'AI-first' than 'we uploaded our API specs'.

Re: WebMCP is available for early preview

#172

Post by charcircuit » Mon, Mar 02, 2026, 10:27 AM UTC

AlecSchueler wrote: post #163

charcircuit wrote: post #129

Earlier quoted context omitted.

Don't use accessibility features either. Just build for humans and let AI understanding take care of understanding all of the details.

Following accessibility best practice is what designing for humans looks like.

The best practices are changing. Many accessibility features were built due to the computer not being understand correctly. For example how something that looks like a checkbox despite being just a div is would not get recognized properly. Now with AI, the AI understands what a checkbox is and can understand from the styling that there is a checkbox there.

Re: WebMCP is available for early preview

#173

Post by thomasfl » Mon, Mar 02, 2026, 10:28 AM UTC

Hywan wrote: post #169

How different is it from the semantic Web (schema, RDF, OWL…)? Instead of reinventing something, why not using a well established technology that can also be beneficial for other usages?

Semantic web is for computers to read data from your website. WebMCP is for interacting with your website.

Using URIs as identifiers and RDF as interchange format, makes it possible for LLM's and computers to understand well what something really means. It makes it well suited for making sure LLM's and computers understand scientific data and are able to aggregate it.

Re: WebMCP is available for early preview

#174

Post by DrScientist » Mon, Mar 02, 2026, 10:34 AM UTC

_heimdall wrote: post #128

TechSquidTV wrote: post #123

Earlier quoted context omitted.

While you absolutely should, I would argue that MCP access would be the OPTIMAL level of accessibility.

Why? What does it add that accessibility features don't cover? And of there's a delta there, why have everyone build WebMCP into their sites rather than improve accessibility specs?

Because, thinking bigger picture, having an AI assistant acting on your behalf might be more effective than slow navigation via accessibility features?

I get the wider point that if accessibility features were good enough at describing the functionality and intent then you wouldn't need a separate WebMCP.

So what does WebMCP do that accessibility doesn't?

Seems to me, at cursory reading, it's around providing a direct js interface to the web site ( as oppose to DOM forms ).

Kind of mixing an API and a human UI into one single page.

Re: WebMCP is available for early preview

#175

Post by DrScientist » Mon, Mar 02, 2026, 10:37 AM UTC

BeefySwain wrote: post #3

Can someone explain what the hell is going on here? Do websites want to prevent automated tooling, as indicated by everyone putting everything behind Cloudfare and CAPTCHAs since forever, or do websites want you to be able to automate things? Because I don't see how you can have both. If I'm using Selenium it's a problem, but if I'm using Claude it's fine??

Obviously if you wanted people to book flights with a bot then you could have provided a public API for that long ago.

I think potentially the subtlety here is a sort of cooperative mode - the computer filling out a lot of the forms and doing the grunt, but it's important that the human is still in the loop - so they need to be able to share a UI with the agent.

Hence a agent friendly web page, rather than just an API.

Re: WebMCP is available for early preview

#176

Post by thrance » Mon, Mar 02, 2026, 10:52 AM UTC

Hywan wrote: post #169

How different is it from the semantic Web (schema, RDF, OWL…)? Instead of reinventing something, why not using a well established technology that can also be beneficial for other usages?

I believe WebMCP will fail for the same reason as the semantic web and public APIs did: no one wants to put in the effort to make their website readable by machines, as that only benefits the competition and is immediately exploited by bad actors.

Re: WebMCP is available for early preview

#177

Post by johneth » Mon, Mar 02, 2026, 11:05 AM UTC

charcircuit wrote: post #172

AlecSchueler wrote: post #163

Earlier quoted context omitted.

Following accessibility best practice is what designing for humans looks like.

The best practices are changing. Many accessibility features were built due to the computer not being understand correctly. For example how something that looks like a checkbox despite being just a div is would not get recognized properly. Now with AI, the AI understands what a checkbox is and can understand from the styling that there is a checkbox there.

Or just use in the first place and save humans and machines a whole bunch of time.

Re: WebMCP is available for early preview

#178

Post by _heimdall » Mon, Mar 02, 2026, 11:48 AM UTC

DrScientist wrote: post #174

_heimdall wrote: post #128

Earlier quoted context omitted.

Why? What does it add that accessibility features don't cover? And of there's a delta there, why have everyone build WebMCP into their sites rather than improve accessibility specs?

Because, thinking bigger picture, having an AI assistant acting on your behalf might be more effective than slow navigation via accessibility features? I get the wider point that if accessibility features were good enough at describing the functionality and intent then you wouldn't need a separate WebMCP. So what does WebMCP do that accessibility doesn't? Seems to me, at cursory reading, it's around providing a direc…

Navigation shouldn't be slow when using accessibility features though. The browser already prices the accessibility tree with full context and semantics of what is on the page and what can be interacted with.

I take the same issue when MCP servers are created for CLI tools. LLMs are very good at running Unix commands - make sure your tool has good `--help` docs and let the LLM figure it out just like a human would.

Re: WebMCP is available for early preview

#179

Post by _heimdall » Mon, Mar 02, 2026, 11:50 AM UTC

charcircuit wrote: post #172

AlecSchueler wrote: post #163

Earlier quoted context omitted.

Following accessibility best practice is what designing for humans looks like.

The best practices are changing. Many accessibility features were built due to the computer not being understand correctly. For example how something that looks like a checkbox despite being just a div is would not get recognized properly. Now with AI, the AI understands what a checkbox is and can understand from the styling that there is a checkbox there.

That's a huge resource cost though, and simply unnecessary. We should be building semantically valid HTML from the beginning rather than leaning on a GPU cluster to parse the function based on the entire HTML, CSS, and JS on the page (or a screenshot requiring image parsing by a word predictor).

Re: WebMCP is available for early preview

#180

Post by me551ah » Mon, Mar 02, 2026, 11:52 AM UTC

_heimdall wrote: post #117

Please don't implement WebMCP on your site. Support a11y / accessibility features instead. If browser or LLM providers care they will build to use existing specs meant to health humans better interact with the web.

Or have an a11y standard for MCPs, where they can't show UI elements and have to only respond with text so that Voice Readers could work out of the box.

This would be a game changer, currenly Voice Readers do not work very well with websites and a11y is a clunky set of tags that you provide to elements and users need to move around elements with back/tab and try to make a mental model of what the website looks like. With MCP and Voice chat, it is like talking to a person.