How do I use mcp-seo?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mcp-seo run a full SEO audit on https://example.com" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

mcp-seo

by g-battaglia

Overview Schema Related Servers Score Discussions

Python

Local

MCP-SEO

Open-source SEO analysis toolkit for AI agents.

MCP-SEO gives any LLM agent — or human — a complete suite of SEO auditing tools, accessible both as a CLI and as a Model Context Protocol (MCP) server. Point it at a URL and get structured, Markdown-formatted reports on meta tags, headings, links, images, performance, mobile-friendliness, structured data, and more.

Features

21 CLI commands covering on-page SEO, technical SEO, performance, accessibility, and content analysis
18 MCP tools exposing the same capabilities to any MCP-compatible client (Claude Desktop, Cursor, OpenCode, ...)
Headless browser rendering via Playwright/Chromium for JavaScript-heavy pages
Lighthouse-style scoring (0-100) with per-category breakdowns
Multi-page site crawler with duplicate detection and robots.txt respect
Structured output — every analyzer returns Pydantic models (JSON) and Markdown reports
Configurable — all timeouts and thresholds overridable via MCP_SEO_* env vars
pip-installable — pip install mcp-seo and you're ready to go

Related MCP server: foglift-mcp

Installation

pip install mcp-seo

# Or with pipx for isolated CLI usage
pipx install mcp-seo

# Or directly from GitHub
pip install git+https://github.com/g-battaglia/mcp-seo.git

After installing, set up the headless browser (one-time):

mcp-seo setup

Quick Start

# Lighthouse-style audit with scoring
mcp-seo lighthouse https://example.com

# Full comprehensive report (all analyzers combined)
mcp-seo report https://example.com

# Individual analyzers
mcp-seo meta https://example.com           # Title, description, OG, Twitter, canonical
mcp-seo headings https://example.com       # Heading hierarchy (h1-h6)
mcp-seo links https://example.com          # Internal/external, nofollow, anchors
mcp-seo images https://example.com         # Alt text, lazy loading, dimensions, format
mcp-seo content https://example.com        # Word count, readability, keywords, n-grams
mcp-seo headers https://example.com        # Caching, security, compression, redirects
mcp-seo sitemap https://example.com        # XML sitemap discovery and validation
mcp-seo robots https://example.com         # robots.txt rules and directives
mcp-seo structured-data https://example.com # JSON-LD, Microdata, RDFa
mcp-seo performance https://example.com    # Core Web Vitals (TTFB, FCP, LCP)
mcp-seo mobile https://example.com         # Viewport, tap targets, font sizes
mcp-seo url-structure https://example.com   # URL length, depth, tracking params
mcp-seo accessibility https://example.com   # ARIA, landmarks, skip-nav, score

# Site crawler
mcp-seo crawl-site https://example.com     # Crawl site, find duplicates

# Page fetching
mcp-seo crawl https://example.com          # Rendered HTML (JS executed)
mcp-seo fetch https://example.com          # Raw HTTP response (JSON)

# Utilities
mcp-seo screenshot https://example.com
mcp-seo og-image -t "Title" -s "Subtitle" -o output.png

MCP Server

MCP-SEO includes a built-in Model Context Protocol server that exposes all SEO tools to any MCP-compatible AI client.

mcp-seo mcp

Claude Desktop

Add to your claude_desktop_config.json:

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

OpenCode

Add to your opencode.json:

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

If you installed from source with uv instead of pip, use "command": "uv" with "args": ["run", "--directory", "/path/to/mcp-seo", "mcp-seo", "mcp"].

Available MCP Tools

Tool	Description
`crawl`	Render page with headless Chromium, return full HTML
`fetch_page`	Fetch raw HTTP response as JSON
`analyze_meta_tags`	Title, description, OG, Twitter, canonical, viewport
`analyze_headings`	Heading hierarchy (h1-h6), single H1, skip checks
`analyze_links`	Internal/external links, nofollow, anchor text
`analyze_images`	Alt text, lazy loading, dimensions, modern formats
`analyze_content`	Word count, readability, keywords, n-grams
`analyze_headers`	Caching, security, compression, redirect chain
`analyze_sitemap`	XML sitemap discovery and validation
`analyze_robots`	robots.txt rules, crawl-delay, sitemap directives
`analyze_structured_data`	JSON-LD, Microdata, RDFa extraction
`analyze_performance`	TTFB, FCP, LCP, DOM nodes, requests, resource sizes
`analyze_mobile`	Viewport, font sizes, tap targets, horizontal scroll
`analyze_url_structure`	URL length, depth, separators, tracking params
`analyze_accessibility`	ARIA landmarks, skip-nav, forms, images, score
`lighthouse_audit`	Lighthouse-style scoring (0-100) per category
`full_seo_report`	Comprehensive report combining all analyses
`crawl_site`	Multi-page crawler with cross-page analysis

For LLM Agents

MCP-SEO is designed to be used autonomously by AI agents. All tools produce structured Markdown output that is easy to parse, quote, and reason about.

See AGENTS.md for recommended workflows: quick audits, comprehensive audits, competitive analysis, and technical audits.

Claude Code Skill

MCP-SEO is available as a Claude Code skill:

npx @anthropic-ai/claude-code skills add g-battaglia/mcp-seo

This installs the mcp-seo skill, giving Claude Code direct access to all 18 SEO analysis tools with recommended audit workflows. The skill auto-triggers on SEO-related prompts and guides the agent to use the right tools for each analysis.

Development

git clone https://github.com/g-battaglia/mcp-seo.git
cd mcp-seo
uv sync --group dev
uv run mcp-seo setup
uv run mcp-seo --help

Tech Stack

Component	Library
CLI	click
MCP server	FastMCP (mcp SDK)
Browser	Playwright (headless Chromium)
HTTP client	httpx
HTML parsing	BeautifulSoup4 + lxml
Data models	Pydantic v2

License

GNU Affero General Public License v3.0

This server cannot be installed

license - permissive license

quality - not tested

maintenance

How are these scores calculated?

Maintenance

–Maintainers

–Response time

–Release cycle

–Releases (12mo)

Commit activity

Resources

GitHub Repository

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Latest Blog Posts

Lightport: Open-Sourcing Glama's AI Gateway
By punkpeye on April 27, 2026.
open source
OpenAI
Tool Definition Quality Score (TDQS)
By punkpeye on April 3, 2026.
mcp
The Hackers Who Tracked My Sleep Cycle
By punkpeye on March 26, 2026.
security

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/g-battaglia/mcp-seo'

If you have feedback or need assistance with the MCP directory API, please join our Discord server