SEO Skill

Name: SEO Skill
Author: ian nuttall

SEO data from your own Search Console, GA4, Bing and site crawls, hosted or run locally over MCP

By ian nuttall (individual) TypeScript Remote (hosted) · Streamable HTTP Local · stdio Data and analytics Browser automation

GitHub npm Website

71 popularity score ★ 609 ↓ 4.3k/wk

Connect your client

Remote (hosted)

Add in one click

Add to Cursor Add to VS Code Add to LM Studio Add to Goose

Opens the app, which asks before it adds the server.

Or add it by command or config

Claude Codeterminal

claude mcp add --transport http seoskill https://seoskill.dev/mcp

Codexterminal

codex mcp add seoskill --url https://seoskill.dev/mcp

Cursor.cursor/mcp.json or ~/.cursor/mcp.json

{
"mcpServers": {
"seoskill": { "url": "https://seoskill.dev/mcp" }
}
}

Claude (desktop and web)Settings, Connectors, add a custom connector

https://seoskill.dev/mcp

Local (npm)

Add in one click

Add to Cursor Add to VS Code Add to LM Studio

Opens the app, which asks before it adds the server.

Or add it by command or config

FirstTerminal

npm i -g seo
seo start

Claude CodeTerminal

claude mcp add seo-skill -- seo mcp serve

Claude Desktopclaude_desktop_config.json

{
  "mcpServers": {
    "seo-skill": {
      "command": "seo",
      "args": [
        "mcp",
        "serve"
      ]
    }
  }
}

Cursor~/.cursor/mcp.json

{
  "mcpServers": {
    "seo-skill": {
      "command": "seo",
      "args": [
        "mcp",
        "serve"
      ]
    }
  }
}

VS Code.vscode/mcp.json

{
  "servers": {
    "seo-skill": {
      "type": "stdio",
      "command": "seo",
      "args": [
        "mcp",
        "serve"
      ]
    }
  }
}

Codex~/.codex/config.toml

[mcp_servers.seo-skill]
command = "seo"
args = ["mcp", "serve"]

Gemini CLITerminal

gemini mcp add --scope user seo-skill seo -- mcp serve

Settings 10

Set together: google-service-account
`SEO_GOOGLE_SERVICE_ACCOUNT_JSON` secret	Optional: Whole Google service account JSON key, used instead of a browser sign-in. Set only one of this, SEO_GOOGLE_SERVICE_ACCOUNT_FILE or GOOGLE_APPLICATION_CREDENTIALS.
`SEO_GOOGLE_SERVICE_ACCOUNT_FILE`	Optional: Path to a service account JSON key file.
`GOOGLE_APPLICATION_CREDENTIALS`	Optional: Standard Google path to a service account key, read when the other two are unset.
Set together: dataforseo
`SEO_DATAFORSEO_LOGIN`	Optional: DataForSEO login. Set together with SEO_DATAFORSEO_PASSWORD. For seo_research.
`SEO_DATAFORSEO_PASSWORD` secret	Optional: DataForSEO API password. For seo_research.
`SEO_AHREFS_API_KEY` secret	Optional: Ahrefs API key for research. For seo_research.
`SEO_BING_API_KEY` secret	Optional: Bing Webmaster Tools API key, used instead of a saved one. For seo_data.
`SEO_CLICKY_SITEKEY` secret	Optional: Clicky sitekey, read instead of asking when connecting a Clicky site.
`TYPESAFE_API_KEY` secret	Optional: TypeSafe key that lets internal link suggestions be classified and judged. For seo_internal_links.
Set together: telemetry
`DO_NOT_TRACK`	Optional: Set to 1 to turn off anonymous usage events. SEO_TELEMETRY_DISABLED=1 does the same.

What it does

SEO Skill gives an assistant the numbers behind a site you own: Search Console clicks, impressions and positions, Google Analytics or Clicky visits, Bing Webmaster rows, dates of Google updates and a crawl of the site. It reports facts and the next tool to call, and leaves the conclusions to the assistant and to you.

There are two ways to use it. The maker runs a hosted server at https://seoskill.dev/mcp that you add to your client and sign in to with Google. It carries the seven shared tools (seo_sites, seo_report, seo_page, seo_compare, seo_index, seo_research and seo_data). The npm package `seo` is a CLI that also starts a local stdio server with `seo mcp serve`, using the Google sign-in and keys saved on your machine.

The README says every command is also an MCP tool. At version 0.13.0 the local server registers 12 tools (packages/mcp/src/index.ts, shared-tools.ts and provider-extension-tools.ts). It has the seven shared ones, plus seo_crawl and seo_internal_links, which only run locally, and three tools for provider packages installed through the CLI. There is no seo_analytics tool. Google Analytics and Clicky rows come through seo_data with its source field. It also serves its usage guide as the resource skill://seo/SKILL.md.

An earlier release exposed a catalogue of about 70 reports behind seo_list_reports and seo_run_report. The maker cut that down to this short set, so older guides that name those tools no longer match.

What it needs

To run it yourself	Node.js 22+
Runs on	macOS, Windows, Linux
Google Search Console	FreeA Google account with access to at least one property. Google Analytics is optional.
DataForSEO or Ahrefs	Paid planOnly for seo_research. Each request uses credits on your own account.
Bing Webmaster Tools	FreeOptional, for Bing rows in the local CLI and server.

Nothing to install for the hosted server: add https://seoskill.dev/mcp and sign in with Google in the browser when the client asks. Codex finishes the sign-in with `codex mcp login seoskill`. For the local server, install the package globally with `npm i -g seo` and run `seo start` once to sign in with Google and add the server to your agents. The package's engines field asks for Node ^22.23.4, ^24.22.1 or 26.12.1 and later. On a machine without a browser, a Google service account JSON key or an SEO_TOKEN made on your seoskill.dev account page replaces the sign-in.

Limits

Not covered Site crawls, internal link suggestions, Bing data and provider packages work only in the local server and CLI, not on the hosted endpoint.
Not covered Traffic, query and index data cover only Search Console properties the signed-in Google account or service account can read.
Cost seo_research calls your own DataForSEO or Ahrefs account and each call spends its credits.
Early release seo_report without a site summarises the weekly site audit, which is in beta. Accounts outside that beta must pass a site.
Size cap seo_internal_links crawls 500 pages by default, up to 5,000 with max_pages.
Rate limit seo_index inspects URLs within the Google URL Inspection quota left for the day.

Without a Search Console property, reports cover crawl evidence only and give no traffic, queries or rankings. Research-provider figures are estimates, and those calls cost money per request (the reports record API units or USD and cache repeats). The crawler identifies itself with a fixed user agent, so sites that block bots may refuse it. There is no hosted endpoint in this release. Reports needing conversion, attribution or geography fields require Google Analytics; Clicky and SerpBase need separate provider packages.

Who it suits

Suits site owners and developers who want an assistant to answer questions from their own Search Console and analytics numbers rather than from estimates. The hosted endpoint needs no install, and the maker's docs say it keeps linked accounts and saved keys but no reports or data rows. The local server suits people who also want crawls, Bing data or everything kept on their own machine. It is a community project, not a Google product. Look elsewhere if you need ranking data for sites you cannot see in Search Console and have no DataForSEO or Ahrefs account.

Tools 12

10 only read, 2 change things. A ? marks an optional parameter.

seo_sites: Lists the Search Console, Google Analytics and Bing properties the connected accounts can read and which research keys are set.
seo_report: Main Search Console report for a site, then one section's rows. With no site, a summary of last week's audited sites.site?: stringsection?: enumstart_date?: stringend_date?: stringdetail?: enumoffset?: integer
seo_page: Live facts for one URL, its Google index status and its Search Console queries against the previous period.url: stringstart_date?: stringend_date?: stringdetail?: enumoffset?: integer
seo_compare: Compares equal periods before and after a date, with the Google updates that fall in the window.site: stringdate: stringpages?: string
seo_index: Google's indexing reasons for known URLs and the pages that lost their clicks, checked live.site: stringstart_date?: stringend_date?: stringmax_urls?: integerdetail?: enumoffset?: integer
seo_research: Keyword, competitor, backlink and AI answer data from your own DataForSEO or Ahrefs key.provider: enumaction: enumkeywords?: arraytarget?: stringcountry?: stringdetail?: enumoffset?: integer
seo_data: Raw rows from Search Console, Google Analytics, Bing or Clicky for questions the report does not cover.source?: enumsite: stringdimensions?: arraypage?: stringquery?: stringstart_date?: stringend_date?: stringdetail?: enumoffset?: integer
seo_crawl changes: Local only: crawls a site from your machine, returns technical issues and saves the crawl.url: string
seo_internal_links: Local only: suggests internal links to add across a site or toward one page.site?: stringurl?: stringmax_pages?: integerstart_date?: stringend_date?: stringdetail?: enumoffset?: integer
seo_list_providers: Local only: lists provider packages installed through the CLI.
seo_describe_provider: Local only: shows one installed provider's capabilities and action schemas.id: string
seo_run_provider changes: Local only: runs one action from an installed provider package.id: stringaction: stringaccount?: objectparams?: objectrefresh?: boolean

Resources 1

skill://seo/SKILL.md: The usage guide for the tools, served as markdown when the package includes it.

Popularity

71 popularity score, on our list (the bar is 36)

★ 609 ↓ 4.3k/wk

GitHub stars	609	33.4 points
Stars in 30 days	+44	8.3 points
npm downloads, last week	4,317	29.1 points

Measured 2026-11-07. How the score works.

Where it is published

Code	iannuttall/seo ★ 609
npm	seo 0.17.1, 2026-11-06
Hosted	`https://seoskill.dev/mcp` Streamable HTTP · OAuth sign-in

What it can reach

Runs	On your machine, or on the maker's servers
Reaches	The internet, Files on your machine, Your account at a service
Changes things	Yes: it can write, send, create or delete
Sign-in	OAuth sign-in
Licence	Apache-2.0 (npm)
Usage data	Sends it unless you turn it off (`DO_NOT_TRACK=1, SEO_TELEMETRY_DISABLED=1 or seo telemetry disable`) Anonymous events such as report started, finished or failed, with client, version, OS and Node major. No URLs or report data. Off when CI is set.
Last code change	2026-11-07

The local server runs on your machine over stdio. It reads Search Console and Google Analytics through a read-only Google sign-in, plus Bing, Clicky and DataForSEO or Ahrefs when keys are set, and fetches the sites you ask it to crawl. Google tokens go to the system keychain, or to a 0600 file in the config directory when no keychain is available. seo_crawl saves crawls locally and seo_run_provider runs code from third-party provider packages, which the CLI installs only after approval in a terminal. The hosted endpoint runs on the maker's servers and, per its docs, stores linked accounts and saved keys encrypted, but no reports or data rows.

Over time

GitHub stars 609 +44 (+8 %)

609565

9 Oct 20267 Nov 2026

npm downloads a week 4.3k +4.2k (+4751 %)

4.3k13

20 Oct 20257 Nov 2026

Before the site measured it, read from npm's daily download counts.

Popularity score 71 +12 (+21 %)

7146

9 Oct 20267 Nov 2026

What is rising across the directory

Sources

Page checked 2026-10-26.

Badge for your README

Shows this server's rank in its category, and stars for a server that runs on your machine. It updates every hour.

MarkdownREADME.md

[![SEO Skill on EveryMCPServer](https://everymcpserver.com/badge/seo-2.svg)](https://everymcpserver.com/servers/seo-2)

Connect your client

Add in one click

Or add it by command or config

Add in one click

Or add it by command or config

Settings 10

What it does

What it needs

Limits

Who it suits

Tools 12

Resources 1

Similar servers

Search Console MCP

OpenSEO MCP

SEOMonster

Google Search Console MCP Server

Browser Use

Bright Data MCP