Google Search Console MCP
MCP server for the Google Search Console API. Enables search performance analytics, sitemap management, URL index inspection, and property management through natural language.
README
Google Search Console MCP
MCP server for the Google Search Console API: search performance analytics (clicks, impressions, CTR, position), sitemap management, URL index inspection and property management — from Claude, Cursor, Codex and other AI clients, in natural language.
Ask your assistant "which queries brought the most clicks last month?", "is this page indexed?", or "resubmit the sitemap" — it drives the Search Console API for you, from a quick indexing check to a full performance analysis.
Quick start
-
Get OAuth credentials for the Search Console API.
-
Add the server — for example, in Claude Code (other clients):
claude mcp add google-search-console \ -e GOOGLE_SEARCH_CONSOLE_CLIENT_ID=your_client_id \ -e GOOGLE_SEARCH_CONSOLE_CLIENT_SECRET=your_client_secret \ -e GOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN=your_refresh_token \ -- npx -y mcp-google-search-console@latest -
Ask the assistant: "Show me the top 20 search queries for example.com over the last 28 days, with clicks and CTR."
Tools
| Tool | Description |
|---|---|
list_sites |
All Search Console properties the account can access, with permission levels. |
get_site |
One property's entry (siteUrl + permissionLevel). |
add_site |
Add a property (stays unverified until verified in Search Console). |
delete_site |
Unlink a property from the account (no data is deleted). |
search_analytics |
Performance query: clicks/impressions/CTR/position by date, query, page, country, device, search appearance or hour; filters, search type, pagination, fresh-data mode. |
get_top_queries |
Shortcut for the most common ask: top queries by clicks, optionally filtered by page substring, country or device. |
list_sitemaps |
Submitted sitemaps (or children of a sitemap index), with errors/warnings. |
get_sitemap |
One sitemap's status and per-content-type counts. |
submit_sitemap |
Submit or resubmit a sitemap. |
delete_sitemap |
Remove a sitemap from Search Console. |
inspect_url |
Google-index status of a URL: verdict, coverage, crawl info, canonicals, rich results. |
raw_request |
Escape hatch: any Search Console API path (webmasters/v3 or v1). |
Plus resilience built in: automatic OAuth token refresh (including on 401), retries with backoff on 429 (and on 5xx/network errors for reads only — mutations are never replayed), a request timeout, and an SSRF guard so the token can't leak to a foreign host.
Example prompts
- "Which pages lost the most clicks this month compared to the previous one?"
- "Is https://example.com/pricing indexed? If not, why?"
- "List queries containing 'mcp' where we rank below position 10."
- "Do any of our sitemaps have errors? Resubmit the ones that do."
Two property formats (the #1 pitfall)
The site_url must match the property exactly as registered in Search Console:
- URL-prefix property — full URL with scheme and trailing slash:
https://example.com/. Notehttp://example.com/,https://example.com/andhttps://www.example.com/are three different properties. - Domain property —
sc-domain:example.com(no scheme, no slash).
A near-match returns 403/404. list_sites shows the exact registered values —
the assistant will usually call it first.
Limitations (the API's, not the server's)
- Dates are in Pacific Time and
end_dateis inclusive; final analytics data lags ~2–3 days (passdata_state: "all"for fresh, still-changing rows). - Filters are AND-only — no OR across filters; regex operators use RE2 syntax.
- URL inspection quota is small: 2,000 inspections per property per day (600/minute) — don't batch-inspect entire sites.
- Search analytics: max 25,000 rows per request (paginate with
start_row); anonymized long-tail queries are never returned. - Verification is out of scope:
add_siteregisters a property, but verification happens in the Search Console UI or the Site Verification API.
Installation
Requires Node.js 20+ (runs via npx, no separate install).
<details open> <summary><b>Claude Code</b></summary>
claude mcp add google-search-console \
-e GOOGLE_SEARCH_CONSOLE_CLIENT_ID=your_client_id \
-e GOOGLE_SEARCH_CONSOLE_CLIENT_SECRET=your_client_secret \
-e GOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN=your_refresh_token \
-- npx -y mcp-google-search-console@latest
</details>
<details> <summary><b>Claude Desktop</b></summary>
claude_desktop_config.json — macOS ~/Library/Application Support/Claude/, Windows %APPDATA%\Claude\
{
"mcpServers": {
"google-search-console": {
"command": "npx",
"args": ["-y", "mcp-google-search-console@latest"],
"env": {
"GOOGLE_SEARCH_CONSOLE_CLIENT_ID": "your_client_id",
"GOOGLE_SEARCH_CONSOLE_CLIENT_SECRET": "your_client_secret",
"GOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN": "your_refresh_token"
}
}
}
}
</details>
<details> <summary><b>Cursor</b></summary>
~/.cursor/mcp.json (or .cursor/mcp.json in the project)
{
"mcpServers": {
"google-search-console": {
"command": "npx",
"args": ["-y", "mcp-google-search-console@latest"],
"env": {
"GOOGLE_SEARCH_CONSOLE_CLIENT_ID": "your_client_id",
"GOOGLE_SEARCH_CONSOLE_CLIENT_SECRET": "your_client_secret",
"GOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN": "your_refresh_token"
}
}
}
}
</details>
<details> <summary><b>VS Code</b></summary>
.vscode/mcp.json — note the servers key (not mcpServers)
{
"servers": {
"google-search-console": {
"type": "stdio",
"command": "npx",
"args": ["-y", "mcp-google-search-console@latest"],
"env": {
"GOOGLE_SEARCH_CONSOLE_CLIENT_ID": "your_client_id",
"GOOGLE_SEARCH_CONSOLE_CLIENT_SECRET": "your_client_secret",
"GOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN": "your_refresh_token"
}
}
}
}
</details>
Getting credentials
The Search Console API has no API-key access — every call needs OAuth 2.0 (the data belongs to your account). One-time setup, ~10 minutes:
- Create a Google Cloud project (or reuse one) at console.cloud.google.com, then enable the Google Search Console API: APIs & Services → Library → Google Search Console API → Enable.
- Configure the OAuth consent screen (APIs & Services → OAuth consent screen): choose External, fill in the app name and your email, and add your Google account under Test users (in Testing mode only listed users can authorize — no app verification needed).
- Create an OAuth client (APIs & Services → Credentials → Create credentials → OAuth client ID), application type Desktop app. Save the client ID and client secret.
- Mint a refresh token. The easiest way is the
OAuth 2.0 Playground:
- Click the gear icon → check Use your own OAuth credentials → paste
your client ID and secret (add
https://developers.google.com/oauthplaygroundas an authorized redirect URI to the OAuth client first). - In Step 1, enter the scope
https://www.googleapis.com/auth/webmastersand click Authorize APIs, signing in with the test-user account that owns your Search Console properties. - In Step 2, click Exchange authorization code for tokens and copy the refresh token.
- Click the gear icon → check Use your own OAuth credentials → paste
your client ID and secret (add
- Put the three values into
GOOGLE_SEARCH_CONSOLE_CLIENT_ID,GOOGLE_SEARCH_CONSOLE_CLIENT_SECRETandGOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN. The server exchanges the refresh token for short-lived access tokens automatically.
Scope notes: https://www.googleapis.com/auth/webmasters covers everything
including sitemap submission and property management. If you only need reads
(analytics, inspection, listings), the narrower
https://www.googleapis.com/auth/webmasters.readonly works — but add_site,
delete_site, submit_sitemap and delete_sitemap will then return 403.
While the consent screen stays in Testing mode, refresh tokens expire after
7 days — publish the app (or keep it Internal in a Workspace domain) for
long-lived tokens.
⚠️ The credentials are stored in plain text in your client's config — treat them like a password. The refresh token grants access to your Search Console data until revoked at myaccount.google.com/permissions.
Configuration
| Variable | Required | Default | Description |
|---|---|---|---|
GOOGLE_SEARCH_CONSOLE_CLIENT_ID |
yes* | — | OAuth2 client id (refresh flow). |
GOOGLE_SEARCH_CONSOLE_CLIENT_SECRET |
yes* | — | OAuth2 client secret (refresh flow). |
GOOGLE_SEARCH_CONSOLE_REFRESH_TOKEN |
yes* | — | OAuth2 refresh token (refresh flow). |
GOOGLE_SEARCH_CONSOLE_ACCESS_TOKEN |
yes* | — | Alternative: a static access token (~1 h lifetime), mostly for testing. |
GOOGLE_SEARCH_CONSOLE_API_BASE |
no | https://searchconsole.googleapis.com |
API root override. |
GOOGLE_SEARCH_CONSOLE_TIMEOUT_MS |
no | 60000 |
Per-request timeout, ms. |
GOOGLE_SEARCH_CONSOLE_MAX_RETRIES |
no | 3 |
Retries on transient errors. |
* Either the three refresh-flow variables together, or GOOGLE_SEARCH_CONSOLE_ACCESS_TOKEN.
Documentation
- All tools — full reference with parameters and notes.
- Development — build, tests, smoke check, telemetry.
- Publishing — releasing and MCP-catalog listing.
Support
Questions, ideas, issues — Telegram @gistrec or GitHub issues.
License
MIT — see LICENSE.
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.