What is WebMCP
Tools for AI agents on web pages
Monitoring components
Components needed for WebMCP
Business benefits
Why monitor WebMCP
Deployment and configuration
Get started
Overview
GermainUX monitors WebMCP activity in real user sessions. It gives teams visibility into how AI agents use the tools a web application exposes, and how those agent actions affect the user’s experience.WebMCP is a proposed web standard. Learn more in the Chrome WebMCP documentation.
What is WebMCP
WebMCP lets a web page expose structured tools to AI agents running in the browser. Instead of simulating clicks and typing on the page (actuation), an agent calls a tool the page has registered, such ascheckout or filter_results.
WebMCP provides:
| Capability | Description |
|---|---|
| Discovery | Pages register the tools that agents can use. |
| JSON Schemas | Each tool defines its expected inputs and outputs. |
| State | The agent and the user share the same page context. |
| API | Description |
|---|---|
| Imperative API | Tools are registered with JavaScript. |
| Declarative API | Tools are defined by annotating HTML forms. |
Monitoring components
| Component | Role |
|---|---|
| GermainUX RUM JS | Monitors WebMCP activity when the GermainUX monitoring script can be deployed in the application. |
| GermainUX RUM Extension | Provides monitoring when RUM JS cannot be deployed directly or access to the complete browser-page context requires an extension. |
| GermainUX JS Profiler | Provides deeper JavaScript, browser, error, and network evidence when a WebMCP problem requires technical investigation. |
Collected data
GermainUX monitors AI agents calling the tools provided within a web page. Each tool execution is collected under the AI Tool Call KPI.Capabilities
GermainUX helps teams:- See which WebMCP tools are used in real user sessions.
- Identify tool calls that fail or return errors.
- Determine which users, sessions, pages, and workflows involve AI agents.
- Compare agent-driven workflows with workflows completed manually.
- Correlate WebMCP activity with user interactions, JavaScript errors, network requests, and Session Replay.
- Use the JS Profiler to investigate the JavaScript and browser activity around a tool call.
Common use cases
WebMCP Monitoring can support applications that expose tools to agents for:- Customer support flows.
- Travel booking.
- Checkout and eCommerce.
- Structured forms and applications.
- Date selection.
- Running diagnostics.
Business benefits
WebMCP Monitoring helps organizations:- Understand how AI agents interact with their web applications.
- Detect failing or unreliable tools before they affect more users.
- Measure the users and business workflows that rely on agents.
- Identify where agents and users struggle to complete a task.
- Prioritize tool improvements according to frequency and business impact.
- Validate whether a correction resolved the problem.
Analytics
GermainUX can analyze WebMCP activity through KPIs, measures, pivots, trends, drill-through views, and session context.| Analysis area | Examples |
|---|---|
| Usage | Tools called, by page and application |
| Reliability | Failed tool calls and recurring errors |
| Impact | Affected users, sessions, pages, and workflows |
| Environment | Application, release, browser, operating system, and device |
| User context | Clicks, navigation, page activity, and Session Replay |
| Technical context | JavaScript errors, long tasks, and related network activity |
Requirements
WebMCP must be available in the user’s browser:- Chrome 149 or later through the WebMCP origin trial.
- For local testing, enable
chrome://flags/#enable-webmcp-testing.
tools Permissions Policy, which defaults to self. Cross-origin iframes need allow="tools" to expose tools.
Limitations
- WebMCP is a proposed standard and may change.
- WebMCP is designed for human-in-the-loop use, not headless agents.
- Agents discover tools only when they visit the site.
- Complex user interfaces may need extra work to expose as tools.
Privacy and security
Tool inputs and outputs may contain sensitive information. Before deployment:- Confirm which applications and tools should be monitored.
- Apply masking and exclusion requirements to Session Replay.
- Restrict access to detailed user and technical evidence.
- Apply appropriate retention and compliance requirements.
- Review the Chrome guidance on secure tools.
Deployment and configuration
- Review Browser Monitoring.
- Review the Browser Monitoring Deployment Overview.
- Deploy GermainUX RUM JS when the monitoring script can be added to the application.
- Deploy the GermainUX RUM Extension when direct script deployment is unavailable.
- Deploy the GermainUX JS Profiler for deeper diagnostics.
- Review the available Browser Monitoring configuration.
Get Help
Contact GermainUX Support
The GermainUX team helps you get up and running faster, empower your team, turn data into actionable insights, and optimize performance to deliver measurable business outcomes.
Can't find what you need?
Search all GermainUX docs with Ctrl + K, or use the feedback buttons at the bottom of this page to tell us what is missing.
The content of this document is proprietary to GERMAIN SOFTWARE.