google-seo-mcp

google-seo-mcp

Provides AI agents with hands-on control of Google SEO and analytics tools including Search Console, GA4, Tag Manager, Indexing API, and PageSpeed Insights, with self-configuring OAuth2.

Category
访问服务器

README

<div align="center">

🔍 google-seo-mcp

Full Google SEO & analytics stack for AI agents — over one MCP server.

Search Console · Indexing API · GA4 (Admin + Data) · Tag Manager · PageSpeed Insights

npm version npm downloads node license MCP

</div>


A Model Context Protocol server that gives Claude (and any MCP client) hands-on control of the Google SEO and analytics stack — list and claim properties, query search performance, submit URLs for indexing, spin up GA4 properties and data streams (i.e. install the measurement tag), wire up Tag Manager, and run PageSpeed audits. Ships with self-configuring OAuth2: no tokens at launch — the agent walks through authorization with built-in tools.

✨ Highlights

  • 🔎 Search Console — properties, sitemaps, search analytics, URL inspection
  • ⚡ Indexing API — push URL_UPDATED / URL_DELETED notifications
  • 📊 GA4 — create properties & web data streams (get the Measurement ID), run reports & realtime
  • 🏷️ Tag Manager — accounts, containers, workspaces, tags
  • 🚦 PageSpeed Insights — Core Web Vitals + Lighthouse, no OAuth needed
  • 🔐 Self-configuring OAuth2 — loopback code capture + auto-refresh, stored locally
  • 🛡️ Safety rails — read-only mode and per-call confirm guards

📦 Installation

Run straight from npm with npx (no install needed), or build from source:

# from source
git clone https://github.com/skiddgoddamn/google-seo-mcp
cd google-seo-mcp
npm install && npm run build

🔌 Connect to your MCP client

Add to your MCP config (e.g. ~/.claude.json or claude_desktop_config.json):

{
  "mcpServers": {
    "google-seo": {
      "command": "npx",
      "args": ["-y", "google-seo-mcp"],
      "env": { "GOOGLE_PSI_API_KEY": "" }
    }
  }
}

<details> <summary>Or point at a local build</summary>

{
  "mcpServers": {
    "google-seo": {
      "command": "node",
      "args": ["F:/projects/google-seo-mcp/build/index.js"]
    }
  }
}

</details>

☁️ Google Cloud setup

  1. Open Google Cloud Console → APIs & Credentials
  2. Create an OAuth 2.0 Client ID → application type Desktop app
  3. Enable these APIs for the project:
    • Google Search Console API · Indexing API · Analytics Admin API · Analytics Data API · Tag Manager API
  4. On the OAuth consent screen, add your Google account as a test user (while the app is in "Testing")
  5. Keep the Client ID and Client Secret handy

🔑 Authorize (self-setup)

No tokens needed at startup — run these tools from your agent, in order:

Step Tool What happens
1 g_set_credentials Save your OAuth Client ID + Client Secret
2 g_authorize Opens Google's consent screen and captures the code via a local loopback server
↳ g_set_code Headless fallback — paste the code + redirect URI manually
✓ g_auth_status Check what's configured and whether the token is valid

The refresh token is saved to ~/.google-seo-mcp/config.json and reused (and auto-refreshed) on every run.

⚙️ Environment variables

Variable Default Description
G_READONLY false Block every mutating tool (_add _create _update _delete _submit _set)
G_CONFIRM false Require a confirm=true argument on mutating calls
G_LOG_LEVEL INFO DEBUG · INFO · WARNING · ERROR
G_LOG_BODIES false Log request/response bodies to stderr (verbose)
G_LOG_FILE — Append log lines to a file (default: stderr only)
GOOGLE_MCP_CONFIG_DIR ~/.google-seo-mcp Override the config directory
GOOGLE_PSI_API_KEY — PageSpeed Insights key; anonymous (rate-limited) if unset

See .env.example for a ready-to-copy template.

🧰 Tool reference

🔐 Auth — g_

Tool Description
g_auth_status Show OAuth configuration status and token validity
g_set_credentials Save OAuth client_id and client_secret
g_authorize Open the consent screen via loopback and store the refresh token
g_set_code Manually exchange an authorization code for a refresh token

🔎 Search Console — gsc_

Tool Description
gsc_sites_get List all Search Console properties
gsc_site_get Get a single property and your permission level
gsc_site_add Add (claim) a property
gsc_site_delete Remove a property
gsc_sitemaps_get List submitted sitemaps
gsc_sitemap_get Get a sitemap's status
gsc_sitemap_submit Submit a sitemap
gsc_sitemap_delete Delete a sitemap
gsc_search_analytics Clicks / impressions / CTR / position by query, page, country, device, date
gsc_url_inspect Index status, coverage, mobile usability, rich results

⚡ Indexing API — idx_

Tool Description
idx_url_submit Notify Google a URL was added/updated (URL_UPDATED)
idx_url_delete Notify Google a URL was removed (URL_DELETED)
idx_metadata Latest indexing-notification metadata for a URL

ℹ️ The Indexing API is officially eligible only for pages with JobPosting or BroadcastEvent structured data. Other URLs may be ignored.

📊 GA4 — ga4_

Tool Description
ga4_accounts_get List Analytics accounts
ga4_properties_get List properties under an account
ga4_property_get Get a property
ga4_property_create Create a GA4 property
ga4_datastreams_get List a property's data streams
ga4_datastream_create Create a web data stream → returns the Measurement ID
ga4_datastream_get Get a data stream
ga4_run_report Run a report (metrics + dimensions over a date range)
ga4_realtime Run a realtime report

🏷️ Tag Manager — gtm_

Tool Description
gtm_accounts_get List GTM accounts
gtm_containers_get List containers under an account
gtm_container_create Create a container → returns its GTM-ID
gtm_workspaces_get List workspaces of a container
gtm_tags_get List tags in a workspace
gtm_tag_create Create a tag in a workspace

🚦 PageSpeed Insights — psi_

Tool Description
psi_run Analyze a URL — Core Web Vitals + Lighthouse (no OAuth required)

🛡️ Safety rails

  • G_READONLY=true — blocks every tool whose name contains _add _create _update _delete _submit _set. Reads still work.
  • G_CONFIRM=true — mutating tools return a preview and require a follow-up call with confirm: true.

📄 License

MIT © skiddgoddamn

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选