WebMCP Readiness Audit -- Get Your Custom Integration Prompt

How the WebMCP Audit Tool Works

The audit tool performs a comprehensive analysis of any website to determine its readiness for WebMCP integration. It evaluates five key areas: forms that could become agent-callable tools, JSON-LD structured data that helps agents understand site content, API endpoints that agents could interact with, meta tags that provide context about the page, and infrastructure signals like sitemap.xml and robots.txt.

1. Enter any website URL into the audit form
2. The tool fetches the page and analyzes its HTML structure
3. You receive a readiness score from 0 to 100 based on detected signals
4. The tool generates suggested WebMCP tool candidates based on your forms and APIs
5. You get a complete, custom prompt with tool definitions and implementation code
6. Paste the prompt into Claude, ChatGPT, Cursor, or any AI coding assistant to build your integration

What the Readiness Score Measures

The readiness score evaluates how prepared a website is for WebMCP integration based on existing signals.

• Forms (up to 25 points): Interactive forms that can be exposed as agent-callable tools, including contact forms, search forms, booking forms, and newsletter signups.
• JSON-LD Structured Data (up to 20 points): Schema.org markup that helps agents understand site content, organization details, and page relationships.
• Open Graph Tags (10 points): Social meta tags that provide agents with page summaries and context.
• API Endpoints (up to 20 points): Detected fetch calls and XHR requests that indicate existing programmatic interfaces.
• Sitemap and Robots.txt (up to 15 points): Infrastructure signals that help agents understand site structure and crawling boundaries.
• Meta Tags (up to 10 points): Descriptive meta information that provides additional context about page content.

API Documentation

The audit tool is available as a public REST API. Send a POST request to /functions/v1/webmcp-audit with a JSON body containing a "url" field and an optional "source" field. The API is open to everyone -- no authentication required.

Request Format

POST with Content-Type: application/json. Required field: url (the website URL to audit). Optional field: source (one of "api", "website", or "webmcp", defaults to "api").

Response Fields

• url: The normalized URL that was audited
• readiness_score: A score from 0 to 100 indicating WebMCP readiness
• webmcp_detected: Whether navigator.modelContext references were found in the page
• findings: Detailed analysis including forms, structured data types, API endpoints, and meta tags
• tool_candidates: Suggested WebMCP tool definitions with names, descriptions, and input schemas
• prompt: A complete implementation prompt ready to paste into an AI coding assistant
• recommendations: Human-readable next steps for improving WebMCP readiness

Rate Limiting

Maximum 10 audits per IP address per hour. Results for the same URL are cached for 24 hours and returned instantly. Rate-limited requests receive HTTP 429 with a retry_after_seconds field.

Our WebMCP Integration -- A Working Reference

This website is itself a WebMCP-enabled site. It registers 7 tools via the navigator.modelContext API that AI agents can discover and call directly when visiting this page in a WebMCP-capable browser.

Read-Only Tools (6)

• get_company_info: Returns structured data about Adapt Marketing including services, location, contact details, and business model.
• get_services: Returns the four types of agentic marketing systems we build with details on capabilities, oversight, and boundaries.
• get_webmcp_services: Returns WebMCP-specific service offerings including audit, schema design, deployment, testing, and optimization.
• get_team_info: Returns team and founder information.
• get_process_overview: Returns the engagement process steps from discovery through deployment.
• run_webmcp_audit: Runs a WebMCP readiness audit on any URL, returning score, findings, tool candidates, and a custom prompt.

Write Tools (1)

• submit_contact_inquiry: Submits a contact inquiry with name, email, company, and message. Writes to the database.

Read-Only vs Write Tool Patterns

Read-only tools return pre-defined structured data and are annotated with readOnlyHint: true so agents know they are safe to call without user confirmation. Write tools perform actions like form submissions and database writes, and agents may ask the user for confirmation before calling them.

How to Verify WebMCP Tool Registration

WebMCP is currently available in Chrome Canary (version 146+) behind an experimental flag. To test tool registration:

1. Install Chrome Canary from google.com/chrome/canary
2. Navigate to chrome://flags/#web-mcp and set it to Enabled
3. Restart Chrome Canary for the flag to take effect
4. Visit this page in Canary
5. Open DevTools console -- you should see "[WebMCP] 7 tools registered successfully."

WebMCP Best Practices

• Write clear, specific descriptions for each tool -- agents rely on these to decide which tool to call
• Use readOnlyHint annotations so agents know which tools are safe to call without confirmation
• Validate all inputs inside execute callbacks before processing
• Return structured JSON data, not raw HTML or plain text
• Use feature detection ('modelContext' in navigator) so sites work normally in browsers without WebMCP
• Keep tool names concise and descriptive using snake_case convention

Need a Full WebMCP Implementation?

The audit tool gives you a starting point. For a full production implementation -- custom tool design, security review, multi-platform testing -- Adapt Marketing can help. We treat WebMCP integration with the same rigor as our agentic marketing systems: you get complete source code, documentation, and training with no recurring fees.

Discuss Full Implementation

Read the WebMCP Deep Dive