Search Console MCP
Provides integration with the Google Search Console API, enabling agents to run code-based searches against the API spec and execute authenticated API requests to manage properties and sitemaps.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Search Console MCPshow my search performance for the last 28 days"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Search Console MCP
A remote MCP server for the Google Search Console API. It runs on Cloudflare Workers and follows Cloudflare's Code Mode pattern (cloudflare/mcp). The agent gets two tools that accept code, instead of one tool per API method.
searchruns the agent's JavaScript against the Search Console API spec, with notes about how the API behaves. The code has no network access.executeruns the agent's JavaScript withgsc.request(), which calls the Search Console API as the logged-in user. It returns the code's result and a call record of every request the code sent.
Users log in with Google and choose read-only or full access. Full access is needed to submit or delete sitemaps and to add or remove properties.
The design decisions are in docs/adr, and the vocabulary is in CONTEXT.md.
Run it locally
You need Node 24 (nvm use reads .nvmrc) and a Google OAuth client. The client must have http://localhost:2530/oauth/callback as a redirect URI.
Install the dependencies with
npm install.Copy
.dev.vars.exampleto.dev.varsand fill inGOOGLE_CLIENT_IDandGOOGLE_CLIENT_SECRET. The file is gitignored.Start the server with
npm run dev. It listens onhttp://localhost:2530.Add
http://localhost:2530/mcpto your MCP client as an HTTP server. The client opens a browser for the Google login.
Related MCP server: Google SearchConsole MCP Server
Tests
npm run check runs the typecheck and the tests. The tests run in Cloudflare's local runtime with the real Worker Loader, and MSW fakes Google.
Two limits are not enforced by the local runtime, so the tests cannot check them. They only take effect on Cloudflare:
the sandbox CPU limit,
the runtime's own request limit. The server enforces its own cap of 50 requests per run, and the tests cover that cap.
Licence
Apache-2.0. Parts are adapted from cloudflare/mcp; see NOTICE.
This server cannot be deployed
Maintenance
Related MCP Connectors
Read-only Search Console analytics, URL inspection, indexing diagnostics, and sitemaps.
Read Search Console performance, keyword opportunities and annotations for your sites.
Google Search Console in your AI: overview, opportunities, index gaps, page checks, long history.
Ask Google Search Console in plain language. Hosted, free, no Google Cloud project.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables interaction with Google Search Console via MCP, offering search analytics, performance summaries, URL inspection, sitemap management, and property listing for SEO workflows.112 npm1MIT
- AlicenseNot gradedqualityDmaintenanceEnables interaction with Google SearchConsole tools and services through a unified API.2MIT
- FlicenseAqualityCmaintenanceEnables automation of Google Search Console tasks, including retrieving search analytics, managing sitemaps and site properties, and inspecting URLs for index status and crawl information.9-
- AlicenseAqualityBmaintenanceEnables read-only access to Google Search Console data, including property discovery, search analytics queries, and URL inspection, through a security-focused interface.326 PyPIMIT