MCP Server

The Nuxt SEO MCP server gives AI assistants the module documentation and free SEO checks for any URL. It needs no account and no API key.

Installation

claude mcp add --transport http nuxt-seo https://nuxtseo.com/mcp

// .cursor/mcp.json
{
  "mcpServers": {
    "nuxt-seo": {
      "type": "http",
      "url": "https://nuxtseo.com/mcp"
    }
  }
}

// .vscode/mcp.json
{
  "mcpServers": {
    "nuxt-seo": {
      "type": "http",
      "url": "https://nuxtseo.com/mcp"
    }
  }
}

// ~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "nuxt-seo": {
      "serverUrl": "https://nuxtseo.com/mcp"
    }
  }
}

// Zed settings
{
  "context_servers": {
    "nuxt-seo": {
      "settings": {
        "url": "https://nuxtseo.com/mcp"
      }
    }
  }
}

1. Go to **Settings** > **Connectors**
2. Add a new connector with URL: `https://nuxtseo.com/mcp`

Tools

Documentation Tools

search_pages: search the full text of the Nuxt SEO docs. Returns the best matching pages with their route, title, and description.

search_pages({ query: 'schema-org breadcrumbs' })

get_page_markdown: read one docs page as Markdown. Use a route from search_pages or list_pages.

get_page_markdown({ route: '/docs/sitemap/getting-started/installation' })

list_pages: list docs pages with their route, title, and description.

list_pages({ limit: 50, offset: 0 })

get_module_setup: get the installation guide for a Nuxt SEO module. Returns the install command, the guide content, and the navigation tree.

get_module_setup({ module: 'schema-org' })

To find content on other pages, use search_pages.

list_modules: list all Nuxt SEO modules with stats (stars, downloads, versions).

list_modules()

SEO Analysis Tools

These tools let AI assistants check and debug the SEO of any public URL.

diagnose_url: run a full SEO diagnosis of a URL in one call. It checks indexability, meta tags, Schema.org, social cards, the sitemap, and robots.txt. It uses the same engine as the free Site Audit tool.

diagnose_url({ url: 'https://example.com' })

Returns:

• score (0 to 100) and grade (A to F)
• summary: one line that states the main result
• findings: ordered by severity, highest first. Each finding has an id, a severity (error, warning, or notice), a category, a title that states the problem, a detail that states the fix, and an optional docs link.
• stats: counts of errors, warnings, notices, and passed checks
• checks: the raw facts that the findings come from
• url, finalUrl (after redirects), status (HTTP status), and fetchedAt

Start with diagnose_url. Use the tools below when you need the full detail of one area.

check_meta_tags: extract and parse the meta tags of a URL.

check_meta_tags({ url: 'https://example.com' })

Returns: parsed title, description, og:*, twitter:*, canonical, robots, and other meta tags.

debug_social_share: debug the social share cards of a URL. Returns the parsed Open Graph and Twitter Card tags with validation warnings.

debug_social_share({ url: 'https://example.com' })

Returns: og:title, og:description, og:image, Twitter Card tags, and warnings for missing or invalid tags.

validate_schema: validate the Schema.org structured data on a URL. Detects JSON-LD and Microdata markup.

validate_schema({ url: 'https://example.com' })

Returns: all detected schemas, validation errors, warnings, and rich result eligibility.

analyze_robots_txt: fetch and parse the robots.txt of a site.

analyze_robots_txt({ url: 'https://example.com' })

Returns: the raw content, the rules for each user agent, and the sitemap directives. If the site has no robots.txt, the result says so. This is not an error: crawlers can then crawl every page.

validate_sitemap: validate the content of an XML sitemap.

validate_sitemap({ xml: '<?xml version="1.0"?>...' })

Returns: validation status, URL count, errors (duplicates, invalid URLs), and warnings.

convert_html_to_markdown: convert the HTML of a page to clean Markdown.

// From URL
convert_html_to_markdown({ url: 'https://example.com' })

// From HTML string
convert_html_to_markdown({ html: '<h1>Hello</h1>', extractContent: true })

Returns: the Markdown content with the original and converted sizes.

Rate Limits

The tools that fetch a URL have the same limits as the free web tools. Each IP address can make 20 requests per minute and 100 requests per day for each tool.

Resources

nuxtseo://reference: a complete reference for all modules, config, and Schema.org patterns in one resource.

URI: nuxtseo://reference

Includes:

• Module overview with key config
• Key composables quick reference
• Complete nuxt.config.ts example
• Schema.org patterns for Article, Product, Person, FAQ, Organization

nuxtseo://modules: a list of all modules with stats (stars, downloads, versions).

URI: nuxtseo://modules

Nuxt SEO Pro MCP

The free server checks one URL at a time. The Nuxt SEO Pro MCP reads the data of the sites you connect. Its tools include:

• next_actions: the ranked list of fixes for a site
• gsc_query: Google Search Console clicks, impressions, and positions
• inspect_url: every recorded issue for one URL, from index coverage, the crawl, Lighthouse, and Core Web Vitals
• crux_vitals: Core Web Vitals field data from the Chrome UX Report
• keyword_research and backlinks: keyword, SERP, and backlink data