# Google Search Console MCP: giving Claude or Cursor your search data

A Google Search Console MCP server is a small program that exposes the Search Console API as tools Claude Code, Cursor or another assistant can call, so the agent can ask which queries a page is shown for while it edits that page. You can run one locally with your own Google Cloud OAuth client, use a hosted connector, or use a product that reads Search Console and exposes its analysis over MCP. Whichever you choose, give it the read-only scope and keep its token out of any file you commit.

Updated 2026-09-23 · Source: https://porteur.ai/guides/google-search-console-mcp

## What the server gives the agent

Without it, a coding agent sees your code and nothing about how Google shows your pages. With it, the agent gets tools that call the Search Console API on your behalf and return rows it can read, sort and compare with the files it is editing.

Most servers wrap two parts of the API. The Search Analytics query method returns clicks, impressions, click-through rate and position, broken down by any of six dimensions: date, query, page, country, device and search appearance. The URL Inspection API returns Google's index status for one URL of a property you have verified.

| Limit | What it means when you ask |
| --- | --- |
| At most 25,000 rows per request | A large site's query and page rows need several requests, paged with startRow. A good server does this for you; check that yours does before trusting a total. |
| Up to sixteen months of data | You can compare seasons, but nothing older. |
| About two to three days of lag | Yesterday's change is not in the data yet. |
| Rare and anonymised queries omitted | Query rows add up to less than the page's totals. That is normal, not a bug in the server. |

Property names trip people up. A domain property is written sc-domain:yourproduct.com. A URL-prefix property is the full URL with a trailing slash, such as https://www.yourproduct.com/. Tell the agent which one you verified, or its first calls will fail with a permission error.

## Three ways to get one

| Route | Who holds your Google token | Built for |
| --- | --- | --- |
| Run an open-source server locally, with your own Google Cloud OAuth client | You, on your machine | Raw figures on demand, full control, some setup |
| A hosted connector | The company that runs it | Quick setup and nothing to run, if you trust the operator |
| A product that reads Search Console and exposes its own analysis over MCP | The product, as with any integration | Findings already worked out, rather than rows the agent has to analyse |

Pick by what you want the agent to do. For "which queries is /pricing shown for", raw rows are enough, and the local route costs nothing but setup time. For "what should I change first", rows are only the start: someone, the agent or a product, has to compare periods, pages and rivals first. As an example of the third route, Porteur exposes a bought report, which includes your Search Console figures when connected, over MCP at https://porteur.ai/mcp with a key from Settings.

Before installing any open-source server, read its code or at least its README and the scope it requests. It will hold a token to your Search Console data.

## Setting up the local route

The steps are the same for every local server, whatever its language. Its README gives the exact command and the variable it reads for credentials.

1. **Create a Google Cloud project** In the Google Cloud console, create a project just for this, such as "search-console-mcp". Keeping it separate makes it easy to delete later.
2. **Enable the Search Console API** In the API library of that project, find the Google Search Console API and enable it.
3. **Configure the consent screen** Set up the OAuth consent screen as an external app in testing, and add your own Google account as a test user. Nobody else needs access.
4. **Create an OAuth client** Create an OAuth client ID of type Desktop app and download its JSON. Save it outside every repository, for example in ~/.config/gsc/, never in the project folder.
5. **Ask for the read-only scope** If the server lets you choose, set the scope to https://www.googleapis.com/auth/webmasters.readonly. An agent that reads data has no reason to hold write access to your properties.
6. **Add it to your agent and sign in** Register the server with Claude Code or Cursor, as below. On the first call it opens a browser window for Google's consent; approve it with the account that has access to the property.

In Claude Code, one command adds the server. Keep it in the default local scope while it holds anything secret:

```bash
# Claude Code, local scope (the default): stored in ~/.claude.json, never committed
claude mcp add gsc -- <command> [args]

# a remote server instead
claude mcp add --transport http gsc <url> --header "Authorization: Bearer <token>"

# check it is connected
claude mcp list
```

Cursor reads MCP servers from .cursor/mcp.json in the project, or ~/.cursor/mcp.json for all projects. Use the file in your home folder when the entry points to credentials:

```json
{
  "mcpServers": {
    "gsc": {
      "command": "<command>",
      "args": ["<args>"],
      "env": {
        "<VARIABLE_THE_SERVER_READS>": "/Users/you/.config/gsc/client_secret.json"
      }
    }
  }
}
```

Test it with one plain request: "list the Search Console properties you can access". If your site is in the answer, the setup works.

## Prompts worth asking

The value is in asking questions the Search Console interface makes slow, and then acting on the answer in the same session. These six cover most of what a small site needs. Paste the block, then work through the answers one page at a time.

```text
Property: sc-domain:yourproduct.com. Period: the last 28 days unless I say otherwise.

1. List the pages with an average position between 4 and 20, sorted by impressions.
   For each, give the three queries with the most impressions and their position.

2. Find queries with many impressions and a click-through rate well below the site's
   average. For each, the page shown and its current title.

3. Compare the last 28 days with the 28 days before. Which pages lost the most
   impressions, and which queries account for the loss?

4. List queries the site is shown for that no page title or h1 in this repository
   mentions. Read the titles from the code, then compare.

5. For /pricing, break down clicks and impressions by device and by country.

6. Inspect https://yourproduct.com/pricing with URL Inspection: what is its index status?

Use only figures returned by the Search Console tools. If a question needs data they
do not return, say so instead of estimating.
```

The first prompt finds pages close to the first page of results, where a clearer title or a better opening paragraph can move them. The second finds titles that are shown but not chosen. The third catches pages sliding before the fall shows in your signups. The fourth is the one only an agent in your repository can answer: it holds the search data and the code side by side, and finds searches you are shown for by accident that deserve a page's title and first paragraph.

Then act on one answer. "The page /guides/csv-import is shown for 'import csv into postgres' at position 12; rewrite its title and first paragraph for that query, and change nothing else" is a task the agent can do and you can judge 28 days later.

## What it cannot answer

Search Console knows your site and nothing else. The server inherits every limit of the data behind it.

- Who else ranks for a query, and what their pages say. Search Console never shows other sites.
- How many people search for something. Impressions count how often your page was shown, not the size of the search.
- Searches you do not appear for at all. If no page of yours is shown, the query is not in your data.
- Anything about rare or anonymised queries, which are left out of the rows.
- The last two or three days, and anything older than sixteen months.

An agent asked one of these questions may still answer from memory. Put a line in your instructions, as in the prompts above, that it must use only figures returned by the tools and say when it has none.

## Keeping it safe

- Use the read-only scope, webmasters.readonly. Full access lets a program submit and delete sitemaps and add or remove properties.
- Keep the OAuth client JSON and the token the server stores outside any repository, and add their folder to .gitignore if it is anywhere near one.
- Never put a token in .mcp.json. Project scope in Claude Code writes that file at the repository root and it is committed. Tokens belong in environment variables or the local scope.
- For a hosted connector, know who runs it, what it stores and how you revoke access. You can remove its access from your Google Account's security settings at any time.
- Remove servers you no longer use. Each one is a live token to your data.

## Questions

### Should I use an export or an MCP server?

An export is enough to start: download the last 28 days, drop the CSV files in a folder the agent can read, and ask your questions. A server is worth the setup once you ask often, or want the agent to compare periods and pages on its own without you exporting again.

### Does a Search Console MCP server work with Cursor?

Yes. Cursor reads MCP servers from .cursor/mcp.json in the project or ~/.cursor/mcp.json in your home folder. The server is the same one Claude Code would use; only the place it is registered changes.

### Can the agent change anything in Search Console?

Only if you grant a write scope. With https://www.googleapis.com/auth/webmasters.readonly it can read performance data and inspect URLs, and nothing more. There is rarely a reason for an agent to hold more.

### Why do the query figures not match the totals?

Search Console leaves rare and anonymised queries out of the query rows, so they add up to less than the page or site totals. The same gap appears in the Search Console interface; the server did not lose data.

### Do I still need the Search Console interface?

Yes, for verification, sitemaps, the indexing reports and anything you want to see as a chart. The MCP server is for questions you want answered inside the session where you change the pages.

## Read next

- [The Search Console API: pull your own search data](https://porteur.ai/guides/search-console-api): What the Google Search Console API gives you, how to authorise it, the request you send, quotas, BigQuery export, and three automations to ship.
- [How to use Google Search Console in ten minutes a week](https://porteur.ai/guides/how-to-use-google-search-console): A quick weekly routine: set four filters, compare 28 days, check pages then queries, and fix three findings, without getting lost in noise.
- [Model Context Protocol (MCP)](https://porteur.ai/glossary/model-context-protocol): The Model Context Protocol in plain words: how MCP servers give Claude or Cursor your data, what that means for SEO work, and what to check before connecting.
- [Claude Code for SEO: what it can fix in your repository, and what it cannot see](https://porteur.ai/guides/claude-code-for-seo): Use Claude Code for SEO: what it fixes in your repo, the search data it lacks, how to bring that data in, a worked session and rules for CLAUDE.md.
- [A Claude SEO skill for your repository: what it should hold, and one to copy](https://porteur.ai/guides/claude-seo-skill): A complete Claude Code SEO skill to paste into .claude/skills: what it checks in your repo, the findings it writes, what it cannot see, and how to adapt it.
- [How to find declining pages in Search Console before the traffic is gone](https://porteur.ai/guides/find-declining-pages-in-search-console): Use Search Console to compare periods, sort by click loss, read queries, and act before a page’s traffic is gone. Works for SaaS and small shops.
- [Striking distance keywords: the searches one page from the light](https://porteur.ai/guides/striking-distance-keywords): Find queries you already rank for at positions 11 to 20, fix the one page that’s close, and add links. Faster wins than writing a new page.
- [Analysing Search Console data with Claude or ChatGPT: the exports and the prompts that work](https://porteur.ai/guides/analyse-search-console-with-ai): Which Search Console export to give ChatGPT or Claude, how to clean it, and five prompts written out whole, each with the table it expects.

Search Console shows your side of each search; the free check adds the rivals on them, reading your site and the searches around it from a URL in about thirty seconds, with three findings whole. Free check: https://porteur.ai/