Repolode
AminForou

AminForou/mcp-gsc

A Model Context Protocol server that connects Google Search Console data to AI assistants like Claude, Cursor and Codex for natural-language SEO analysis.

Stars
1,984
Forks
276
Open issues
15
Last commit
last month

What is it?

mcp-gsc is a Model Context Protocol (MCP) server that links Google Search Console to AI assistants. Once configured in a client such as Claude Desktop, Cursor, Codex CLI or Gemini CLI, you can ask questions about your site's search performance in plain language, such as which queries drive traffic or which pages dropped in clicks.

The server exposes about 20 tools covering property management, search analytics with filters by country, device, query and page, period-over-period comparison, URL inspection for indexing status, batch URL checks of up to 10 URLs, and sitemap listing and submission. Authentication works with either an OAuth desktop client that uses your own Google account or a service account for automation and team use. Destructive operations such as deleting sites or sitemaps are disabled unless explicitly enabled.

The server runs locally over stdio by default, and it can also be deployed over SSE with Docker for remote use.

Key features

  • 01Query search analytics by country, device, query and page with filters and sorting.
  • 02Compare performance between two date ranges.
  • 03Inspect URL indexing status, including rich-result issues, in single or batch form.
  • 04List, inspect and submit sitemaps.
  • 05Supports OAuth for personal accounts and service accounts for automation.
  • 06Destructive tools are disabled by default.

Use cases

  • Asking an AI assistant for a weekly SEO performance summary of a site.
  • Finding queries where several pages compete and deciding which to keep.
  • Auditing indexing problems across a set of important pages.
  • Identifying high-impression, low-CTR queries to improve titles.

Quick start

  1. 1
  2. 2

    Create Google API credentials (OAuth client or service account) and save the JSON file

  3. 3

    Add the uvx server config to your AI client (e.g. Claude Desktop), pointing to your credentials file

  4. 4

    Restart the client and ask it to list your GSC properties

Things to know

  • Requires setting up Google Cloud credentials and enabling the Search Console API before it works.
  • A hosted paid version with limited seats is promoted in the README.

Topics

FAQ

What is mcp-gsc?
A Model Context Protocol server that connects Google Search Console data to AI assistants like Claude, Cursor and Codex for natural-language SEO analysis. mcp-gsc is a Model Context Protocol (MCP) server that links Google Search Console to AI assistants. Once configured in a client such as Claude Desktop, Cursor, Codex CLI or Gemini CLI, you can ask questions about your site's search performance in plain language, such as which queries drive traffic or which pages dropped in clicks.
Is mcp-gsc free to use? What license does it have?
mcp-gsc is open source under the MIT license. Check the license terms before using it in commercial products.
Is mcp-gsc still maintained?
Updated within 6 months. The last commit to mcp-gsc was on Sep 15, 2026.
What can I use mcp-gsc for?
Asking an AI assistant for a weekly SEO performance summary of a site. Finding queries where several pages compete and deciding which to keep. Auditing indexing problems across a set of important pages. Identifying high-impression, low-CTR queries to improve titles.
How do I install mcp-gsc?
Install uv (run the installer) Create Google API credentials (OAuth client or service account) and save the JSON file Add the uvx server config to your AI client (e.g. Claude Desktop), pointing to your credentials file Restart the client and ask it to list your GSC properties
Where can I get mcp-gsc?
The source code and installation instructions are on GitHub: https://github.com/AminForou/mcp-gsc

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects