ownership.manifestRead only

codeComplete source code repository, yours
docsArchitecture docs, operations docs, training
deployYour AWS, Azure, GCP or on-premise
dataAdapt never holds client data
termsMonthly retainer, no long-term contract
oversightAgents draft and suggest; humans approve

How we build. Not live telemetry.

Free WebMCP Readiness Audit v2

Audit, Grade &
Improve Your WebMCP

Deep analysis of imperative tools, declarative forms, annotations, schemas, and agent-invoked patterns. Get a letter grade, score breakdown, and an AI-ready prompt to build or improve your integration.

Instant analysis
Detects both APIs
Score breakdown
No signup required

Try the Audit Tool

This audit covers the act layer. For the read layer, run the free Accessibility Tree Audit.

WebMCP Readiness Audit

Enter any website URL below. We analyze forms, structured data, APIs, meta tags, declarative tools, and existing WebMCP integrations -- then generate a custom prompt to build or improve your integration.

Our WebMCP Integration

This site registers 8 tools via navigator.modelContext -- 7 read-only and 1 write. AI agents visiting this page can discover and call these tools directly.

read-only

`get_company_info`

Returns structured data about Adapt Marketing: services, location, contact, business model.

read-only

`get_services`

Returns the four agentic marketing system types with details on capabilities, oversight, and boundaries.

read-only

`get_webmcp_services`

Returns WebMCP-specific service offerings: audit, schema design, deployment, testing, optimization.

read-only

`get_team_info`

Returns team and founder information.

read-only

`get_process_overview`

Returns the engagement process steps from discovery through deployment.

write

`submit_contact_inquiry`

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

read-only

`run_webmcp_audit`

Runs a WebMCP readiness audit on any URL. Returns score, findings, tool candidates, and a custom prompt.

read-only

`run_accessibility_tree_audit`

Runs an agent readability audit on any URL. Returns a 0 to 100 score, findings by category, recommendations, and a fix prompt. Read only.

Read-Only Pattern

Read-only tools return pre-defined structured data. They don't hit external APIs or write to databases. Set readOnlyHint: true in annotations so agents know these are safe to call without confirmation.

annotations: { readOnlyHint: true }

Write Pattern

Write tools perform actions -- submitting forms, writing to databases. They validate inputs inside the execute callback and return success/failure status. Agents may ask the user for confirmation before calling these.

annotations: { readOnlyHint: false }

The Audit Tool Bridges Everything

The run_webmcp_audit tool is callable three ways: from this page's interactive form, by AI agents via WebMCP tool registration, and through the REST API. All three hit the same edge function, share the same rate limits, and return the same structured results with a generated prompt.

API Documentation

The same endpoint powers the on-page audit tool, the WebMCP-registered tool, and external API calls. Open to all, no auth required. v2 adds grading, score breakdown, tool quality analysis, and improvement suggestions.

Endpoint

POST /api/audits/webmcp

Request Body

url required The website URL to audit (e.g., "https://example.com")
source optional Where the request originates: "api", "website", or "webmcp". Defaults to "api".

Response Fields

url

Normalized URL that was audited

readiness_score

0-100 readiness score

grade

Letter grade: A, B, C, D, or F

webmcp_detected

Whether navigator.modelContext was found

webmcp_tools_detail

Per-tool name, description, annotations, schema

findings

Forms, structured data, APIs, declarative tools

tool_quality

Quality metrics: descriptions, annotations, schemas

score_breakdown

Points per category showing how score was earned

score_improvements

Ranked actions to improve score with point values

tool_candidates

Suggested WebMCP tool definitions

prompt

Complete implementation or improvement prompt

recommendations

Human-readable next steps

duration_ms

Audit execution time in milliseconds

cached

True if served from 24h cache

Examples

cURL

bash

curl -X POST \
  https://www.adaptmarketing.com/api/audits/webmcp \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com", "source": "api"}'

JavaScript

javascript

const response = await fetch(
  'https://www.adaptmarketing.com/api/audits/webmcp',
  {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      url: 'https://example.com',
      source: 'api'
    })
  }
);

const audit = await response.json();
console.log(audit.grade);  // "B"
console.log(audit.score_breakdown);
console.log(audit.prompt);

Response

json

{
  "url": "https://example.com",
  "readiness_score": 45,
  "grade": "C",
  "webmcp_detected": false,
  "webmcp_tool_count": 0,
  "webmcp_detection_source": "none",
  "findings": {
    "forms": [...],
    "structured_data": ["Organization", "WebSite"],
    "api_endpoints": ["/api/search"],
    "meta_tags": [...],
    "sitemap_exists": true,
    "robots_exists": true,
    "has_json_ld": true,
    "has_open_graph": true,
    "declarative_tools": [],
    "agent_invoked_detected": false
  },
  "tool_candidates": [...],
  "tool_quality": {
    "total": 0, "withDescriptions": 0,
    "withAnnotations": 0, "withInputSchema": 0,
    "readOnly": 0, "mutation": 0, "categories": []
  },
  "score_breakdown": {
    "webmcp_base": 0, "webmcp_tools": 0,
    "declarative_tools": 0, "annotations": 0,
    "input_schemas": 0, "forms": 8,
    "structured_data": 12, "open_graph": 5,
    "api_endpoints": 5, "sitemap": 3,
    "robots": 3, "meta_tags": 4,
    "agent_invoked": 0, "total": 40
  },
  "score_improvements": [
    { "action": "Implement WebMCP tool registration", "points": 35 },
    { "action": "Add JSON-LD structured data", "points": 12 }
  ],
  "prompt": "# WebMCP Integration Prompt for...",
  "recommendations": [...],
  "duration_ms": 2340
}

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.

How to Verify

Test WebMCP tool registration in Chrome Canary with the experimental flag enabled.

Testing in Chrome Canary

Chrome 146+ Canary includes experimental WebMCP support

01 Install Chrome Canary
Download from google.com/chrome/canary if you don't have it.
02 Enable the WebMCP flag
Navigate to chrome://flags/#web-mcp and set it to Enabled.
03 Restart Chrome Canary
The browser needs a full restart for the flag to take effect.
04 Visit this site
Open https://www.adaptmarketing.com/webmcp-integration in Canary.
05 Open DevTools console
You should see "[WebMCP] 8 tools registered successfully." logged.

Expected Console Output

[WebMCP] Registering tools...
[WebMCP] 8 tools registered successfully.

Best Practices

Write clear, specific descriptions -- 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 degrade gracefully
Keep tool names concise and descriptive using snake_case convention

Ready to Make Your Site AI-Agent Ready?

The audit tool gives you a starting point. For a full production implementation -- custom tool design, security review, multi-platform testing -- we can help.

Discuss Full Implementation Read the WebMCP Deep Dive

No signup, no sales pitch. The audit tool is free and open to everyone.