Deeprank MCP

Deeprank MCP brings methodology and optimization guidance from deeprank.org into your IDE. When you ask to build an AI-optimized site or audit your project, the AI gets our checklist, file-level hints, and schema suggestions—no tab-switching, no guesswork.

Without Deeprank MCP

LLMs rely on generic or outdated advice about AI search:

  • Generic “add meta tags” instead of file-level steps
  • No alignment with the Deeprank protocol (identity, exclusions, llms.txt)
  • No stack-specific hints (e.g. where to add schema in Next.js)

With Deeprank MCP

The AI pulls methodology and tools from our MCP—directly into your prompt.

  • Concrete steps: e.g. “Add wrong_choice_when to Product additionalProperty in layout.tsx”
  • Same 16 checks we use, with “where to look” for your stack
  • Copy-paste JSON-LD and llms.txt structure

Install

Add the Deeprank MCP server in your IDE or agent. No repo clone—just add the config below.

Production (use this in your IDE or agent):

{
  "mcpServers": {
    "deeprank": {
      "url": "https://mcp.deeprank.ai/mcp"
    }
  }
}

If you already have mcpServers, add only the deeprank entry:

"deeprank": {
  "url": "https://mcp.deeprank.ai/mcp"
}

Cursor

  1. Open Settings → Features → MCP.
  2. Click Add new MCP server → choose URL (Streamable HTTP).
  3. Or edit .cursor/mcp.json and add the deeprank block from above.
  4. Save. No API key required (optional for higher limits).

All MCP clients

Installation examples for popular MCP clients. Deeprank MCP is remote only (URL): https://mcp.deeprank.ai/mcp. Optional: add headers with Authorization: Bearer YOUR_API_KEY for higher limits.

Cursor▼

Paste into ~/.cursor/mcp.json or project .cursor/mcp.json. Or: Settings → Cursor Settings → MCP → Add new global MCP server → URL → https://mcp.deeprank.ai/mcp.

{
  "mcpServers": {
    "deeprank": {
      "url": "https://mcp.deeprank.ai/mcp"
    }
  }
}
Claude Code▼

Remote server (Streamable HTTP). See Claude Code MCP docs.

claude mcp add --transport http deeprank https://mcp.deeprank.ai/mcp
Windsurf▼

Add to your Windsurf MCP config. Windsurf MCP docs.

{
  "mcpServers": {
    "deeprank": {
      "serverUrl": "https://mcp.deeprank.ai/mcp"
    }
  }
}
Cline▼

MCP Servers → Remote Servers → Edit Configuration. Cline MCP.

{
  "mcpServers": {
    "deeprank": {
      "url": "https://mcp.deeprank.ai/mcp",
      "type": "streamableHttp"
    }
  }
}
VS Code (Copilot MCP)▼

Add to your VS Code MCP config. VS Code MCP docs.

"mcp": {
  "servers": {
    "deeprank": {
      "type": "http",
      "url": "https://mcp.deeprank.ai/mcp"
    }
  }
}
Claude Desktop▼

Settings → Connectors → Add Custom Connector. Name: Deeprank, URL: https://mcp.deeprank.ai/mcp.

Other clients (generic)▼

Most MCP clients that support Streamable HTTP accept this. Add the deeprank entry to your mcpServers or equivalent.

{
  "mcpServers": {
    "deeprank": {
      "url": "https://mcp.deeprank.ai/mcp"
    }
  }
}

What the MCP provides

  • Optimization steps — What to do and in what order, aligned with deeprank.org.
  • Audit checklist — The same 16 checks we use (8 basic + 8 AI enhanced), with “where to look” hints for your stack (Next.js, WordPress).
  • Exclusion examples — “When not to recommend” by industry, and copy-paste JSON-LD for wrong_choice_when and identity.
  • llms.txt structure — Recommended sections and examples.

What you can ask

Try prompts like:

  • “I want to build an AI optimized site.”
  • “I want to create a project that is optimized for AI search visibility.”
  • “Audit my codebase for AI readiness.”
  • “How do I add wrong_choice_when to my schema?”
  • “What’s the recommended llms.txt structure?”

The AI will use the Deeprank MCP to pull in the right methodology and give file-level, stack-aware suggestions.

Free & rate limits

The MCP is free for everyone with a reasonable rate limit. Optional: add an API key from your Deeprank account in the server config (e.g. header Authorization: Bearer YOUR_API_KEY) for higher limits.

Next steps

Getting started with DeeprankPricing & plansMCP endpoint