01
WebMCP API signals
Looks for public code that references the proposed browser APIs and model context surface.
Free WebMCP readiness checker
See what browser-agent tools a public page exposes.
Check public code for WebMCP signals, named tools, descriptions, schemas, handlers, annotations, discovery files, and fallback handling.
WebMCP in one sentence
WebMCP is a proposed browser standard that lets a webpage expose structured tools so a supported agent can request page actions without relying only on screenshots and simulated clicks.
Static public-code detection only. The checker never executes detected tools and does not guarantee runtime compatibility.
Free previewStatic public check
Check if a website publishes WebMCP signals
Run a free public preview. Account holders unlock the complete signal check, tool explanations, communication flow, and starter recommendations.
Five free checks per hour. Public pages only. The Lab never executes detected tools, submits forms, or reads private accounts.
What the scanner checks
Eight signals of a clearer WebMCP implementation
API detection alone is not enough. Useful browser-agent tools need clear intent, structured inputs, reliable handlers, sensible capability signals, fallbacks, and human control.
02
Registered tool names
Finds structured tool names that communicate the actions a supported agent may request.
03
Tool descriptions
Reviews descriptions that help an agent distinguish one page action from another.
04
Input schemas
Checks for structured input definitions, required values, types, and field guidance.
05
Executable handlers
Reports public evidence that tool definitions include action handlers without executing them.
06
Annotations and capability
Checks signals that may describe read-only and write-capable behavior.
07
Discovery and fallback
Reviews public discovery files, feature detection, and non-WebMCP fallback handling.
08
Human control
Highlights the need for visible review, permission, validation, and confirmation around sensitive actions.
MCP made clear
WebMCP complements MCP. It does not replace it.
The simplest distinction is the operating context. MCP connects an AI application with an external server. WebMCP exposes structured tools through the webpage currently open in a supported browser.
| Question | MCP | WebMCP |
|---|---|---|
| Operating context | AI client and external server | Open webpage and browser context |
| Typical capabilities | Tools, resources, prompts, data, and services | Page tools, forms, UI logic, and current page state |
| Open website tab required | Not necessarily | Yes for the intended browser workflow |
| Primary use | Backend and external-system integration | Visible human-agent collaboration on the page |
Common questions
WebMCP FAQ
What is WebMCP?+
WebMCP is a proposed web standard that lets a webpage expose structured tools to supported AI agents through browser APIs and annotated page controls. It is designed for interaction inside an open webpage with the user still involved.
Is WebMCP the same as Model Context Protocol?+
No. Model Context Protocol connects an AI application with an external MCP server. WebMCP exposes page-level tools inside the browser. The two approaches can complement each other.
Does this scanner execute detected tools?+
No. The public checker performs static public-code detection. It reports visible implementation signals and never calls the tools found on the scanned website.
Does a detected tool prove runtime support?+
No. Static detection cannot prove that a compatible browser registered the tool successfully, that permissions are granted, or that an agent can complete the action. Runtime testing is a separate step.
Does WebMCP improve search rankings?+
WebMCP is an interaction layer, not a search ranking factor. Search visibility still depends on crawlability, indexing, useful content, internal linking, page quality, and other search signals.
Can every browser agent use WebMCP?+
No. Support depends on current browser and client implementations, permissions, origin rules, and the evolving proposal. Websites should keep accessible human controls and reliable fallbacks.