Hybrid Ranking MCP Server

What is MCP?

The Model Context Protocol (MCP) lets AI assistants call external tools natively. Once installed, your AI assistant can look up Hybrid Ranking checks — what we audit and how to fix each issue — directly inside any conversation, with current data instead of training-cutoff snapshots.

Endpoint

JSON-RPC over HTTP POST. No authentication is required in this MVP — the free tier exposes the public check catalog.

https://hybridranking.com/mcp

Claude Desktop setup

Add the following entry to your claude_desktop_config.json under mcpServers, then restart Claude Desktop:

{
  "mcpServers": {
    "hybridranking": {
      "type": "http",
      "url": "https://hybridranking.com/mcp"
    }
  }
}

Available tools

• list_checks — list the public Hybrid Ranking check catalog. Optional filters: category, tier, page_types.
• get_check — fetch the full detail for one check (display name, what we check, how to fix, sources, canonical URL). Required: check_key.

Rate limits

• 60 tools/call invocations per IP per hour (the heavier path).
• 200 discovery requests per IP per hour (tools/list, resources/list, prompts/list).
• Returned as HTTP 429 with a Retry-After hint when exceeded.
• If you need higher limits for production use, contact us — a Pro tier with authenticated, per-user limits ships in Phase 2.

Browse the catalog

Prefer the web view? See the full check catalog at hybridranking.com/checks. Companion deep-dive: see the Hybrid Ranking blog.