seo-stack-mcp

seo-stack-mcp

The complete open-source SEO data stack for AI agents, integrating Google Search Console, Google Analytics 4, Bing Webmaster Tools, and Microsoft Clarity with built-in analyst tools like cannibalization check, low-hanging fruit, and content gap analysis, all self-hosted and free.

Category
访问服务器

README

seo-stack-mcp

<!-- mcp-name: io.github.Raffaele86/seo-stack-mcp -->

The complete open-source SEO data stack for AI agents. Google Search Console, Google Analytics 4, Bing Webmaster Tools and Microsoft Clarity in a single MCP server — with built-in analyst tools (cannibalization check, low-hanging fruit, content gap, weekly report), not just raw API wrappers.

Self-hosted, free, MIT. Works with Claude Desktop, Claude Code, Cursor, and any MCP client.

Why this instead of N separate servers?

  • One install, one config — instead of wiring up 4 different MCP servers with 4 auth setups.
  • Analyst tools included — ask "which queries are declining this month?" or "find my low-hanging-fruit keywords" and get an answer computed server-side, not a raw data dump your agent has to crunch.
  • Sources are optional — configure only what you use; tools appear only for configured sources.
  • Self-hosted — your tokens stay on your machine. No third-party hosted gateway reading your Search Console.

Quickstart (60 seconds)

Add to your MCP client config (Claude Desktop claude_desktop_config.json, Cursor mcp.json, or claude mcp add):

{
  "mcpServers": {
    "seo-stack": {
      "command": "uvx",
      "args": ["seo-stack-mcp"],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/service-account.json",
        "GA4_PROPERTY_ID": "123456789",
        "BING_WEBMASTER_API_KEY": "your-key",
        "CLARITY_API_TOKEN": "your-token"
      }
    }
  }
}

Every env var is optional — set only the sources you want. That's it: restart your client and ask "what are my top queries this week?".

To run the latest development version straight from GitHub, use "args": ["--from", "git+https://github.com/Raffaele86/seo-stack-mcp", "seo-stack-mcp"] instead.

Credentials

Source What you need Where to get it
GSC + GA4 Service account JSON (recommended) → GOOGLE_APPLICATION_CREDENTIALS Google Cloud Console → create service account → enable Search Console API + Analytics Data API → add the SA email as a user in GSC / GA4
GSC + GA4 (alt) OAuth desktop client JSON → SEO_STACK_OAUTH_CLIENT First run opens a browser for consent; refresh token cached in ~/.config/seo-stack-mcp/
Bing BING_WEBMASTER_API_KEY Bing Webmaster Tools → Settings → API access
Clarity CLARITY_API_TOKEN Microsoft Clarity → project → Settings → Data Export API

Optional defaults: GSC_SITE_URL, BING_SITE_URL (so you don't repeat the site on every question).

Tools

Google Search Console (29 tools)

Tool Description
gsc_bulk_inspect_urls Inspect multiple URLs in batch via the URL Inspection API. Max 50 URLs per call.
gsc_bulk_request_indexing Request indexing for multiple URLs via the Google Indexing API. Max 50 URLs.
gsc_cannibalization_check Find queries where multiple pages compete (keyword cannibalization). Shows queries with 2+ ranking pages.
gsc_compare_periods Compare two periods: shows deltas of clicks, impressions, CTR and position per item.
gsc_content_gap_analysis Content gap analysis: queries where the competitor ranks but you don't (or you are far behind).
gsc_country_performance Performance by country: clicks, impressions, CTR and average position. Default: last 28 days.
gsc_daily_stats Aggregated daily site statistics: clicks, impressions, CTR and average position per day.
gsc_declining_queries Queries losing clicks/positions: compares last N days vs the equivalent previous period.
gsc_delete_sitemap Remove a sitemap from Google Search Console. Requires write scope on the credentials.
gsc_device_performance Compare performance across MOBILE vs DESKTOP vs TABLET. Default: last 28 days.
gsc_indexing_status_summary Indexing status summary for multiple URLs: how many indexed, not indexed, and errored.
gsc_inspect_url Inspect a URL via the Google Search Console URL Inspection API.
gsc_keyword_opportunities Queries with high impressions but low CTR in position 5-20: optimization opportunities.
gsc_list_sitemaps List all sitemaps of a site in Google Search Console.
gsc_list_sites List all Google Search Console properties available to the configured credentials.
gsc_low_hanging_fruit Queries in position 3-10 with high impression volume: small optimizations can push them into the top 3.
gsc_page_trend Daily trend for a specific page over the last N days (default 90).
gsc_pages_for_query Which pages rank for a specific query? Shows all pages with performance data.
gsc_queries_for_page Which queries drive traffic to a specific page? Shows all keywords with performance data.
gsc_query_trend Daily trend for a specific query over the last N days (default 90).
gsc_request_indexing Request indexing of a URL via the Google Indexing API.
gsc_rising_queries Rising queries: compares last N days vs the equivalent previous period.
gsc_search_analytics Query Search Analytics: clicks, impressions, CTR and position for the requested dimensions.
gsc_search_analytics_filtered Search Analytics with advanced filters. Filters: query_contains, query_regex, page_contains, page_regex, country (e.g. 'ita'), device (MOBILE/DESKTOP/TABLET).
gsc_search_appearance Performance by SERP result type (rich snippet, video, FAQ, etc.). Default: last 28 days.
gsc_submit_sitemap Submit a sitemap to Google Search Console. Requires write scope on the credentials.
gsc_top_pages Top pages by clicks or impressions. order_by: clicks, impressions, ctr, position. Default: last 28 days.
gsc_top_queries Top queries by clicks or impressions. order_by: clicks, impressions, ctr, position. Default: last 28 days.
gsc_weekly_report Automatic weekly report: last 7 days vs previous 7 days metrics with deltas and top queries/pages.

Google Analytics 4 (10 tools)

Tool Description
ga4_compare_periods Compare two periods with percentage delta on key metrics.
ga4_conversion_paths User paths: acquisition source → landing page with sessions and users.
ga4_events Top GA4 events with count and number of users.
ga4_landing_pages Entry pages (landing pages) with sessions, users and bounce rate.
ga4_overview Site overview: users, sessions, pageviews, bounce rate, average session duration.
ga4_page_performance Detailed performance of a single page with device and source breakdown.
ga4_realtime Realtime report: pages currently being viewed and active users.
ga4_top_pages Top pages by traffic with full metrics (pageviews, users, duration, bounce rate).
ga4_traffic_sources Traffic breakdown by source/medium (e.g. google/organic, direct/none).
ga4_user_demographics User demographics breakdown: country, device, browser, language.

Bing Webmaster Tools (22 tools)

Tool Description
bing_crawl_issues URLs with crawl problems (4xx/5xx errors, redirects, etc.).
bing_crawl_stats Bing crawl statistics: crawled pages, errors, redirects, etc.
bing_keyword Bing impression volume for an exact query (impressions + broad).
bing_keyword_stats Historical statistics for a keyword on Bing: impressions and trend.
bing_link_counts Pages with the highest number of backlinks according to Bing.
bing_pages_for_query Pages that rank for a specific query on Bing.
bing_queries_for_page Search queries that drive traffic to a specific page.
bing_query_page_detail Detailed statistics for a specific query + page combination.
bing_query_traffic Daily traffic trend for a specific query on Bing.
bing_related_keywords Keywords related to a search term on Bing, with impression data.
bing_sitemaps List all sitemaps/feeds submitted to Bing.
bing_sites List all sites verified in the Bing Webmaster Tools account.
bing_submission_quota Check the URL submission quota for Bing.
bing_submit_sitemap Submit a new sitemap/feed to Bing.
bing_submit_url Submit a URL to Bing for indexing.
bing_submit_urls_batch Submit a batch of URLs to Bing for indexing (max 500).
bing_top_pages Top pages on Bing with clicks, impressions and average position.
bing_top_queries Top Bing search queries with clicks, impressions and average position.
bing_traffic_stats Daily Bing traffic statistics: clicks and impressions over time.
bing_url_info Indexing status of a specific URL on Bing.
bing_url_links Inbound backlinks to a specific URL according to Bing.
bing_url_traffic Traffic data for a specific URL on Bing.

Microsoft Clarity (10 tools)

Tool Description
clarity_breakdown Free-form raw breakdown (1-3 dimensions). Power-user — costs 1 quota unit if not already cached.
clarity_dead_clicks Top URLs with dead clicks (clicks on non-interactive elements — a UX problem signal).
clarity_engagement Engagement Time + Scroll Depth per URL (or another dimension).
clarity_excessive_scroll Top URLs with excessive scrolling (users hunting for info they can't easily find).
clarity_popular_pages Top URLs by visits over the last days (Popular Pages metric, dimension=URL).
clarity_quickback_clicks Top URLs with quickback clicks (user goes back immediately — page not relevant).
clarity_quota_status Daily quota status (local counter, does not call the API).
clarity_rage_clicks Top URLs with rage clicks (fast repeated clicks — user frustration).
clarity_script_errors Top URLs with JS errors + error clicks (technical problems on the page).
clarity_traffic Traffic breakdown from Microsoft Clarity (sessions, bot %, pages per session).

Why not the official GA4 server or mcp-gsc?

Both are good and this project doesn't pretend otherwise. Google's official analytics-mcp covers GA4 only; mcp-gsc covers GSC only. If you live in one data source, use them. This project is for people who work across the whole free SEO stack and want the cross-source workflow (GSC + GA4 + Bing + Clarity) plus opinionated analysis tools in one place.

License

MIT.


Built and maintained by Raffaele Nocera — freelance web & SEO. Available for consulting.

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