Search DevTools

Jump to any tool or page

SearchConsole.ai

Read-only Search Console analytics, URL inspection, indexing diagnostics, and sitemaps.

AndreyAzimov0 stars0 forksAI & Agents
View source

Install

mcp_config.json

{
  "mcpServers": {
    "ai-searchconsole-searchconsole-ai": {
      "url": "https://searchconsole.ai/mcp",
      "type": "streamable-http"
    }
  }
}

Documentation

Google Search Console MCP (GSC) — SearchConsole.ai

SearchConsole.ai is a hosted, read-only Google Search Console MCP server for GSC search analytics, SEO performance, URL inspection, indexing diagnostics, and sitemaps. Use the Google Search Console API as an SEO MCP integration in Codex, Claude Code, Cursor, ChatGPT, or another OAuth-capable MCP client without creating a Google Cloud project.

npx -y searchconsole-ai@latest setup

Or choose a client directly:

npx -y searchconsole-ai@latest setup codex
npx -y searchconsole-ai@latest setup claude
npx -y searchconsole-ai@latest setup cursor

Google Search Console MCP features

  • No Google Cloud project or downloaded OAuth credentials
  • Hosted Streamable HTTP MCP endpoint
  • Google OAuth with read-only Google Search Console access
  • GSC search analytics, period comparisons, URL inspection, indexing diagnostics, and sitemaps
  • Works with Codex, Claude Code, Cursor, ChatGPT, and other OAuth-capable MCP clients

What the setup command does

For Codex and Claude Code, the CLI invokes the client's supported mcp add command and starts its MCP OAuth flow. For Cursor, it safely adds this entry to ~/.cursor/mcp.json, preserving existing configuration and creating a backup before changing an existing file:

{
  "mcpServers": {
    "searchconsole-ai": {
      "type": "http",
      "url": "https://searchconsole.ai/mcp"
    }
  }
}

Preview changes without applying them:

npx -y searchconsole-ai@latest setup cursor --dry-run

Configure a command-line client without opening OAuth immediately:

npx -y searchconsole-ai@latest setup codex --no-login

This package does not contain or run a local Search Console server. It never asks for, receives, or stores Google credentials. Authentication is handled by the MCP client and the hosted SearchConsole.ai OAuth service.

Manual configuration

Add a remote Streamable HTTP MCP server named searchconsole-ai with this URL:

https://searchconsole.ai/mcp

Full instructions: searchconsole.ai

Public source and release history: github.com/AndreyAzimov/searchconsole-ai

How to use it

After configuring the server, complete the client's OAuth sign-in with a Google account that has access to your Search Console properties. Then ask your AI assistant to use SearchConsole.ai, specifying the property and date range you want to analyze.

Example prompts:

  • "For example.com, compare clicks, impressions, CTR, and average position for the last 28 complete days with the preceding 28 days. Which pages lost the most clicks?"
  • "Find queries for example.com with high impressions and low CTR over the last 28 complete days. Suggest improvements based on the Search Console data."
  • "Inspect https://example.com/my-page and explain the indexing status reported by Google Search Console."
  • "List the sitemaps for example.com and summarize any reported errors or warnings."

Replace the example domain and URL with your own. SearchConsole.ai reads Search Console data; it does not edit your website, submit sitemaps, or request indexing.

Privacy and support

License

MIT

Sourced from the repository README.

More in AI & Agents