MCP Ecommerce

Installation guide

How to install mcp-server-gsc (Google Search Console)

In Claude, ChatGPT, Gemini, Microsoft Copilot, Perplexity, Cursor and Claude Code, with what to check before you start and the first prompts to run.

This is a local server. You run the `mcp-server-gsc` npm package on your own computer through `npx`, and it authenticates to Google with a service account JSON key referenced by the `GOOGLE_APPLICATION_CREDENTIALS` environment variable. There is no hosted URL and no browser sign-in. The server exposes one tool, `search_analytics`, and is read-only.

Have someone with Google Cloud project access and Search Console owner rights do the setup, ideally an SEO lead or agency developer. Use a dedicated service account, not a personal login. Keep the JSON key file private, because anyone holding it can read the data of every property the service account was added to. The project is community maintained, so review the repository before adopting it.

mcp-server-gsc (Google Search Console) setup guide: Community's MCP documentation page on github.com, shown in a browser frame (seo MCP server for e-commerce)
mcp-server-gsc (Google Search Console) setup guide preview: Community’s documentation on github.com, captured Oct 13, 2026. The page may have changed since; the link opens the live version.

Before you start

  • Node.js 18 or later on the computer that will run the server.
  • A Google Cloud project with the Search Console API enabled (APIs & Services > Library).
  • A service account with a JSON key downloaded (APIs & Services > Credentials), stored in a private location, with its absolute path noted.
  • Permission in Search Console to add the service account email (name@project.iam.gserviceaccount.com) to each property as a property administrator, as the docs describe.
  • The exact site URL of each property: a full URL such as https://example.com/ or sc-domain:example.com for domain properties.
How it runs
Local, runs on your machine
Access
Read-only
Server address
Copy it from the vendor documentation

mcp-server-gsc (Google Search Console) in Claude

  1. Complete the Google Cloud and Search Console setup first: enable the Search Console API, create a service account, download the JSON key, and add the service account email to each property as a property administrator.
  2. Open the Claude Desktop configuration file (claude_desktop_config.json), reachable from the Settings developer section. The docs do not give the file location, so use the standard one for your operating system.
  3. Add the gsc entry under mcpServers as shown in the snippet, and replace /path/to/credentials.json with the absolute path to your key file.
  4. Save the file and fully restart Claude Desktop.
  5. Check that the gsc server and its search_analytics tool are listed, then run a small test query.
claude_desktop_config.json
{
  "mcpServers": {
    "gsc": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-server-gsc"
      ],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/credentials.json"
      }
    }
  }
}

mcp-server-gsc (Google Search Console) in ChatGPT

  1. The documentation does not cover ChatGPT. This server runs locally over stdio, and ChatGPT custom connectors in Developer mode expect a remote MCP server URL.
  2. The repository mentions a remote server for Cloudflare Workers in its commit history, but the README gives no URL or setup steps for it. Do not guess a URL.
  3. If you want to use it with ChatGPT, you would need to host it yourself behind a remote MCP endpoint. Take the exact method from the vendor documentation or the repository source.
  4. For a supported path, use a client that can launch local commands, such as Claude Desktop, Cursor or Claude Code.

mcp-server-gsc (Google Search Console) in Gemini

  1. The documentation does not cover Gemini. The standard approach for Gemini CLI is to add a local server under mcpServers in its settings JSON file.
  2. Complete the Google Cloud and Search Console setup first, and keep the JSON key path ready.
  3. Add an mcpServers entry named gsc with command set to npx, args set to ["-y", "mcp-server-gsc"], and env containing GOOGLE_APPLICATION_CREDENTIALS.
  4. Restart Gemini CLI and confirm the gsc server is listed.
  5. Run a small test query for one site and a short date range.
Gemini CLI settings (mcpServers)
{
  "mcpServers": {
    "gsc": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-server-gsc"
      ],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/credentials.json"
      }
    }
  }
}

mcp-server-gsc (Google Search Console) in Microsoft Copilot

  1. The documentation does not cover Microsoft Copilot Studio or GitHub Copilot. The steps below use the standard VS Code format for a local server.
  2. Complete the Google Cloud and Search Console setup first and note the absolute path to your JSON key.
  3. In your workspace, create or open .vscode/mcp.json and add a gsc entry under servers as shown in the snippet.
  4. Replace /path/to/credentials.json with your key path, save, and start the server from the MCP tools list in VS Code.
  5. Copilot Studio uses remote connectors and this server has no documented hosted URL, so it is not a documented fit.
.vscode/mcp.json
{
  "servers": {
    "gsc": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-server-gsc"
      ],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/credentials.json"
      }
    }
  }
}

mcp-server-gsc (Google Search Console) in Perplexity

  1. The documentation does not cover Perplexity.
  2. In the Perplexity desktop app, open the connectors or MCP settings and look for an option to add a local server by command. Availability depends on your app version and plan.
  3. If offered, enter npx as the command, -y mcp-server-gsc as the arguments, and set GOOGLE_APPLICATION_CREDENTIALS to the absolute path of your JSON key.
  4. If the app only accepts remote URLs, this server is not a fit, because the documentation gives no hosted endpoint.
  5. Run a small test query after saving.

mcp-server-gsc (Google Search Console) in Cursor

  1. Complete the Google Cloud and Search Console setup first and note the absolute path to your JSON key.
  2. Open Cursor MCP settings, or edit the global ~/.cursor/mcp.json or a project .cursor/mcp.json. The docs do not give these paths, so confirm them in Cursor's documentation.
  3. Add the gsc entry under mcpServers as shown in the snippet.
  4. Replace /path/to/credentials.json with your key path and save.
  5. Enable the server in Cursor's MCP settings and confirm search_analytics appears.
mcp.json
{
  "mcpServers": {
    "gsc": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-server-gsc"
      ],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/credentials.json"
      }
    }
  }
}

mcp-server-gsc (Google Search Console) in Claude Code

  1. Complete the Google Cloud and Search Console setup first and note the absolute path to your JSON key.
  2. Run the claude mcp add command from the snippet in your terminal. The --env flag is a standard Claude Code option, not from the server docs.
  3. Replace /path/to/credentials.json with your key path.
  4. Run claude mcp list to confirm gsc is registered, or use /mcp inside a session.
  5. Start with a small test query for one site.
Terminal
claude mcp add gsc --env GOOGLE_APPLICATION_CREDENTIALS=/path/to/credentials.json -- npx -y mcp-server-gsc

Check that it works

Start with questions that only read data. If these answer from your real account, the connection is good.

  • Using the gsc server, run search_analytics for siteUrl `sc-domain:example.com` from 2024-01-01 to 2024-01-07 with dimensions `query` and rowLimit 10.
  • Show the top 10 pages by clicks for `https://example.com/` over the last 7 days, using dimensions `page` and rowLimit 10.
  • Run search_analytics for one site over a short date range with detectQuickWins enabled and the default quickWinsConfig, and list the first 10 results.

Troubleshooting

+Why do I get a permission or 403 error for a site?

The service account probably has not been added to that property. In Search Console, add the service account email (`name@project.iam.gserviceaccount.com`) as a property administrator, as the docs say, and repeat for each property. Also check that the Search Console API is enabled in the same Google Cloud project as the key.

+Why does the gsc server not appear in my client?

Check that the JSON is valid and the entry sits under `mcpServers` (or `servers` in VS Code). Confirm Node.js 18 or later and `npx` are on the PATH the client uses, then fully restart the client. Make sure `GOOGLE_APPLICATION_CREDENTIALS` points to an existing absolute file path.

+Why does a query return no data or an invalid site error?

Use the site URL exactly as it is registered in Search Console. The docs give `http://www.example.com/` style full URLs or `sc-domain:example.com` for domain properties. Dates must be `YYYY-MM-DD`. Very recent dates may have little data, and `dataState` defaults to `final`, so try `all` if you need fresher figures.

Written Oct 13, 2026 from the vendor documentation. Exact menus, addresses and commands change; the official documentation is the source of truth.