Google Search Console MCP Server

Install the Google Search Console MCP server via npm to use it with MCP-compatible clients.

npm install mcp-server-gsc

To use this server, you must set up a Google Cloud Project and a Service Account with access to your Search Console properties.

1. Enable the API: In the Google Cloud Console, go to "APIs & Services" > "Library", search for "Search Console API", and enable it.
2. Create a Service Account:
   • Navigate to "APIs & Services" > "Credentials".
   • Click "Create Credentials" > "Service Account".
   • Complete the setup and create a new key in JSON format. This file will be downloaded automatically.
3. Grant Property Access:
   • Open your Google Search Console property.
   • Add the service account email (e.g., name@project.iam.gserviceaccount.com) as a Property Administrator.

The enhanced_search_analytics tool allows for more complex queries than the standard search_analytics tool. It supports:

• High Row Limits: Up to 25,000 rows.
• Regex Filters: Advanced filtering using regular expressions.
• Quick Wins Detection: Identifying optimization opportunities based on specific thresholds.
• Dimension Filtering: Filtering by page, query, country, or device.

Note that country and device filters use an equals operator, while page and query filters can use a user-provided filterOperator (e.g., for regex).

Add the following configuration to your Claude Desktop configuration file to enable the gsc MCP server. Ensure the GOOGLE_APPLICATION_CREDENTIALS environment variable points to the absolute path of your downloaded JSON credentials file.

{
  "mcpServers": {
    "gsc": {
      "command": "npx",
      "args": ["-y", "mcp-server-gsc"],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/credentials.json"
      }
    }
  }
}

The detect_quick_wins tool automates the process of finding SEO 'low-hanging fruit'. It performs a comprehensive analysis by fetching up to 25,000 rows of search data and then applying thresholds to identify queries or pages that are performing well but have room for improvement.

It uses the following thresholds for detection:

• minImpressions: Minimum impressions required.
• maxCtr: Maximum Click-Through Rate (CTR) to target.
• positionRangeMin: Minimum search position.
• positionRangeMax: Maximum search position.

The tool returns a JSON object containing the list of quickWins, the totalOpportunities count, and the applied thresholds.

The server is designed to run as a Node.js executable using stdio transport. To run the server, you must provide a path to your Google Service Account credentials via the GOOGLE_APPLICATION_CREDENTIALS environment variable. The server will then start and listen for requests over standard input/output.

export GOOGLE_APPLICATION_CREDENTIALS="/path/to/your/service-account-file.json"
node path/to/mcp-server-gsc/dist/index.js

Below are advanced usage patterns for the search_analytics tool, including regex filtering and Quick Wins detection.

// Advanced Filtering with Regex
{
  "siteUrl": "https://example.com",
  "startDate": "2024-01-01",
  "endDate": "2024-01-31",
  "dimensions": "page,query",
  "queryFilter": "regex:(AI|machine learning|ML)",
  "filterOperator": "includingRegex",
  "deviceFilter": "MOBILE",
  "rowLimit": 10000
}

// Quick Wins Detection
{
  "siteUrl": "https://example.com",
  "startDate": "2024-01-01",
  "endDate": "2024-01-31",
  "dimensions": "query,page",
  "detectQuickWins": true,
  "quickWinsConfig": {
    "positionRange": [4, 15],
    "minImpressions": 500,
    "minCtr": 2
  }
}

The createServer function is the primary entrypoint for initializing a Google Search Console MCP server. It requires a credentials string (typically a service account key or authentication token) to initialize the SearchConsoleService. The returned Server instance provides access to various Google Search Console tools via the Model Context Protocol.

Usage

import { createServer } from './src/server.js';

const credentials = 'YOUR_GOOGLE_SEARCH_CONSOLE_CREDENTIALS';
const server = createServer(credentials);

// The server instance can now be used with an MCP transport (e.g., Stdio)

The service provides methods to manage sitemaps for a site. Like searchAnalytics, these methods automatically attempt to normalize the siteUrl to a sc-domain: format if a permission error occurs.

• listSitemaps(requestBody): Lists sitemaps for a site.
• getSitemap(requestBody): Retrieves a specific sitemap.
• submitSitemap(requestBody): Submits a new sitemap.

All methods accept request bodies conforming to the webmasters_v3 API types.

// List sitemaps
const sitemaps = await service.listSitemaps({ siteUrl: 'https://example.com/' });

// Submit a sitemap
await service.submitSitemap({
  siteUrl: 'https://example.com/',
  feedpath: '/sitemap.xml'
});