I get the excitement, but let’s be honest about what we actually have right now.

This is metadata duct taped onto the UI and a system that relies entirely on browser security. And really — do you trust that? Most people already run a handful of third party extensions without thinking twice.

WebMCP isn’t a standard. It’s not even a draft. It’s an early preview with:

• missing discovery mechanisms

• unclear permission models

• no cross browser consensus

• no defined security model for malicious agents

• no rate limiting or throttling

• no guidance for sensitive actions

• no validation

• can be spoofed

• can be misconfigured

It feels very much like the early days of:

• schema.org (inconsistent adoption)

• AMP (half baked, then abandoned)

• Web Components v0 (rewritten later)

And we all remember what happened when early AMP adopters dominated rich results… right up until the whole thing collapsed.

I’m not saying the direction is wrong — just that the current implementation is nowhere near ready for the kind of trust people are already projecting onto it.

Feels less like an open standard and more like Chrome trying to secure its place in an agent driven web before someone else defines the rules.