gsc-mcp
Provides read-only access to Google Search Console data, including search analytics, sitemap status, and URL inspection, for MCP clients like Claude.
README
gsc-mcp
A free, open-source Model Context Protocol server that gives Claude (or any other MCP client -- ChatGPT, Codex, Cursor) direct, read-only access to your Google Search Console data: search analytics (clicks/impressions/CTR/position), sitemap status, and URL inspection.
Built by GrowthLever because, as a marketer, I don't love the interface for Google's products -- Search Console, GA4, and Ads included. This lets me put my own translation layer on top of the raw data and ask Claude for the insight directly, instead of digging through Google's UI. It's free because the tool isn't the product -- the workflows I run on top of it are (see the bottom of this page).
Bring your own Google credentials
This runs entirely on your machine using a Google OAuth client you create in your own Google Cloud account. Your refresh token never leaves your disk, and no Search Console data ever passes through a server GrowthLever operates. There's no account to create with us, nothing to sign up for, nothing to trust us with. See SECURITY.md and PRIVACY.md for the full detail.
What it exposes
Four read-only tools:
| Tool | What it does |
|---|---|
list_sites |
Lists the GSC properties the authorized account can access |
search_analytics |
Clicks/impressions/CTR/position, filterable by query, page, country, device, date |
list_sitemaps |
Submitted sitemaps and their status/errors |
inspect_url |
Index coverage, canonical URL, and crawl details for one URL |
Nothing here can change Search Console settings or submit/delete anything -- every call is a GET or a read-only query/inspect request.
Get started
QUICKSTART.md -- 10-minute setup, most of it waiting on Google's console. Short version:
npx -y @growthlever/gsc-mcp auth --credentials ~/Downloads/client_secret.json
npx -y @growthlever/gsc-mcp doctor
Then point your Claude config at npx -y @growthlever/gsc-mcp (example in examples/claude-desktop-config.json) and ask it something -- see PROMPTS.md for a starter pack, split by founder-building-from-zero and scale-up-optimizing-what-works.
Something not working? Run gsc-mcp doctor first, then check TROUBLESHOOTING.md.
Local development
npm install
node bin/gsc-mcp.js auth --credentials <path to client_secret.json>
node bin/gsc-mcp.js doctor
node bin/gsc-mcp.js # runs the MCP server over stdio
node test/smoke.js # spawns the server and verifies tool registration
Docs
- QUICKSTART.md -- setup, start to finish
- PROMPTS.md -- starter prompt pack
- SECURITY.md -- what this touches, what it can't do, how to revoke access
- PRIVACY.md -- what's collected (nothing) and who sees your data
- TROUBLESHOOTING.md -- plain-language fixes for common Google/OAuth errors
- examples/ -- config for Claude Desktop and Codex CLI, plus a weekly-review prompt template
Why this is free
I'm a fractional Head of Growth -- most of the founders and operators I talk to are starting from scratch and didn't get into business to learn marketing. The best early move is almost always laying an organic foundation while referrals and word of mouth do the rest, because there isn't budget for paid yet. Scale-ups have a different problem: figuring out which pages and content are actually working so they can double down instead of guessing.
Either way, the blocker usually isn't access to data -- Search Console is free -- it's translating that data into "what do I do next." That's what this tool plus a decent prompt does. I'm giving the tool away because the tool isn't hard to build; the judgment about what to do with the output is where the actual value is, and that's what I write about and go deeper on for subscribers.
If you want the applied layer -- prompt packs by use case, weekly review and client-reporting templates, and the founder/scale-up workflows this is built around -- that's on Substack (subscriber content).
License
MIT. Fork it, self-host it, build on it.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.