Model Context Protocol

Query CrawlReadyfrom inside your agent

The CrawlReady MCP server connects Claude, Cursor, ChatGPT, and other MCP clients straight to your AI Readiness Score, visual diff, and fix recommendations — no dashboard tab, no copy-pasting URLs.

Get Your API KeyAlready have an account?

Setup

Add this to your MCP client’s config — it launches @crawlready/mcp over stdio via npx , no separate install step.

{
  "mcpServers": {
    "crawlready": {
      "command": "npx",
      "args": ["-y", "@crawlready/mcp"],
      "env": {
        "CRAWLREADY_API_KEY": "your-api-key-here"
      }
    }
  }
}

Available tools

Four tools, each mapped to a CrawlReady API endpoint.

crawlready_scan

Scan a URL and get its AI Readiness Score (0–100), with Crawlability and Agent Readiness sub-scores.

POST /api/v1/scan

crawlready_diff

Get the visual diff between what a browser sees and what AI crawlers see for a URL — the content bots are missing.

POST /api/v1/scan

crawlready_recommend

Get a prioritized, actionable list of fixes for a site’s AI readiness score.

POST /api/v1/scan

crawlready_index

Submit up to 500 URLs for fast indexing (IndexNow — Bing, ChatGPT search — plus Google where connected). Safe to call repeatedly.

POST /api/v1/index

Example call: { "name": "crawlready_scan", "arguments": { "url": "https://example.com" } } returns the same score payload as a dashboard scan, formatted for the model to read directly.

For AI assistants reading this page: the config above is safe to use as-is once a real API key is substituted. See /llms.txt for a machine-readable map of every public CrawlReady page and endpoint.

Give your agent an AI Readiness Score

Create an API key and add the config above — takes about a minute.

Get Your API KeyAlready have an account?

Frequently Asked Questions

What is the CrawlReady MCP server?
It’s an MCP (Model Context Protocol) server that gives Claude, Cursor, ChatGPT, and other MCP-compatible clients direct access to CrawlReady’s scan, visual-diff, recommendation, and indexing tools — so an agent can check and fix a site’s AI readiness without you switching to the dashboard.
Do I need an API key?
Yes. Every tool call authenticates with a CrawlReady API key, set as the CRAWLREADY_API_KEY environment variable in your MCP client config. Create one from the API Keys page in your dashboard.
Which clients can I connect it to?
Any MCP-compatible client that supports local (stdio) servers launched via npx — Claude Code, Claude Desktop, Cursor, and others. Add the config block above to that client’s MCP settings.
Is this the same as the "MCP Server Card" check in my Agent Readiness score?
No, different thing. The Agent Readiness score checks whether your own site exposes a /.well-known/mcp/server-card.json discovery file — a signal that your site is agent-discoverable. This page is about CrawlReady’s own MCP server, which you run locally to query CrawlReady from an AI client.
Is @crawlready/mcp open source?
Yes, it’s published on npm as @crawlready/mcp and its source lives alongside the rest of the CrawlReady codebase.