search-console

The Search Console MCP is a unified AI-driven gateway to Google Search Console, Bing Webmaster Tools, and Google Analytics 4. It lets you query and analyze ranking, visibility, and engagement data across multiple engines and accounts from a single context window. You can use the built-in prompts and workflows to detect traffic anomalies, identify opportunities (such as high-impression pages with low engagement), check for keyword cannibalization, compare performance across engines, and generate prioritized action plans. The server handles authentication across accounts, resolves credentials per site, and returns structured insights that you can feed directly to your AI agent for decision making.

To get started, install and run the MCP, then use the supplied prompts or your own natural-language requests to query data like top keywords, cannibalization issues, opportunity scores, or cross-engine performance. The server also offers operational prompts for data health checks and ROI-focused analyses, so you can move beyond raw CSV exports to actionable intelligence. You can manage multiple Google and Bing accounts, attach sites to accounts, and let the server automatically select the appropriate credentials during queries.

Prerequisites:

• Node.js >= 14 (or a compatible runtime that supports npx)
• npm or pnpm

Step-by-step installation:

1. Ensure Node.js and npm are installed

node -v
npm -v

2. Install and run the MCP server via npx (no global install required)

npx -y search-console-mcp

3. First-run setup (authentication):

• For interactive login (Desktop Flow), run the setup command:

npx search-console-mcp setup

• This will start a local secure server, open the browser for Google authorization, and store credentials in your keychain.

4. For service account (advanced) workflows, configure Google credentials via environment variable(s) as described in the readme, e.g.:

export GOOGLE_APPLICATION_CREDENTIALS="/path/to/your/key.json"

5. Optional: configure Bing API key

export BING_API_KEY="your-api-key-here"

6. Start the MCP server and begin issuing prompts from your MCP client. The exact startup command may vary slightly depending on how you deploy, but the recommended approach is the npx-based workflow shown above.

Tips and common issues:

• Ensure you have the correct API keys and credentials for Google Search Console, Bing Webmaster Tools, and GA4 if you plan to query across all engines.
• For server-side automation, prefer the Service Account flow for GA4 integration and set GOOGLE_APPLICATION_CREDENTIALS accordingly.
• The MCP stores tokens securely (OS keychains / credentials managers). If you switch machines, you may need to re-authenticate.
• If you need to run the server in a container or CI, use the npx-based invocation or package the MCP runner as specified by the project guidelines and ensure environment variables are securely provided.
• Multi-account management commands help you list, add, or remove accounts and attach sites to accounts; use these to keep credentials organized as you scale to 20+ accounts.