Rampify

Rampify MCP Server lets you run an MCP (Model Context Protocol) server that brings SEO intelligence into your editor. It enables real-time SEO audits, Google Search Console insights, and AI-powered recommendations directly in your workflow, so you can validate and optimize pages without leaving your IDE or terminal.

How to use

You interact with the Rampify MCP Server through your MCP client. Install the server, configure your API key and domain, then run the MCP server alongside your editor so Claude Code or Cursor can invoke tools like get_page_seo, get_issues, crawl_site, and more. Use natural language prompts to trigger the right tool, or call a specific tool by name if you know it.

How to install

Prerequisites you need before installation:

• Node.js 18 or higher

• Rampify account (free to start)

Install via npm

npm install -g @rampify/mcp-server

Manual setup and per-project configuration

Configure the MCP server for a project so each project can point to its domain. You will provide your API key and the project domain in the environment for the MCP server.

Per-project setup (example shown for Claude CLI):

cd /path/to/your/project
claude mcp add -s local -t stdio \
  --env BACKEND_API_URL=https://www.rampify.dev \
  --env API_KEY=sk_live_your_api_key_here \
  --env SEO_CLIENT_DOMAIN=your-domain.com \
  rampify -- npx -y @rampify/mcp-server

# Reload your IDE window

Global setup for Claude CLI

If you want global access across projects (domain must be specified per request):

claude mcp add --scope user rampify npx \
  -y @rampify/mcp-server \
  --env BACKEND_API_URL=https://www.rampify.dev \
  --env API_KEY=sk_live_your_api_key_here

# Reload your IDE window

Manual configuration (Cursor)

Add to your Cursor settings or ~/.cursor/config.json the Rampify MCP server configuration.

{
  "mcpServers": {
    "rampify": {
      "command": "npx",
      "args": [
        "-y",
        "@rampify/mcp-server"
      ],
      "env": {
        "BACKEND_API_URL": "https://www.rampify.dev",
        "API_KEY": "sk_live_your_api_key_here",
        "SEO_CLIENT_DOMAIN": "your-domain.com"
      }
    }
  }
}

Manual configuration (Claude Code)

{
  "mcpServers": {
    "rampify": {
      "command": "npx",
      "args": [
        "-y",
        "@rampify/mcp-server"
      ],
      "env": {
        "BACKEND_API_URL": "https://www.rampify.dev",
        "API_KEY": "sk_live_your_api_key_here",
        "SEO_CLIENT_DOMAIN": "your-domain.com"
      }
    }
  }
}

Environment variables to know

The server uses a few environment variables to configure its behavior. You typically provide BACKEND_API_URL, API_KEY, and an optional SEO_CLIENT_DOMAIN at runtime.

Optional defaults include CACHE_TTL and LOG_LEVEL to control caching behavior and log verbosity.

Configuration options at a glance

Key environment variables you will configure for successful operation.

• BACKEND_API_URL (required): Rampify API endpoint, typically https://www.rampify.dev

• API_KEY (required): Your Rampify API key that starts with sk_live_

• SEO_CLIENT_DOMAIN (optional): Default domain for the current project

Troubleshooting

If you encounter issues, verify that the backend is reachable, your API key is valid, and the domain is correctly set for your project. Check logs by setting LOG_LEVEL=debug in your environment.

Notes and best practices

• Run the per-project setup to keep your domain context consistent across tasks.

• Use get_issues for site-wide audits and crawl_site after deployments to refresh data.

Development

To develop locally, install dependencies, run the backend, and start the MCP server in tandem with your editor.

Available tools

get_page_seo

Get comprehensive SEO data for a specific page, including issues, domain defaults, and optional local vs production data.

get_issues

Fetch site-wide SEO issues with a health score and prioritized recommendations.

get_gsc_insights

Retrieve Google Search Console performance data plus AI-powered content recommendations.

generate_meta

Create optimized meta tags (title, description, OG/Twitter tags) with optional client profile context.

generate_schema

Auto-generate and validate schema.org JSON-LD for a page, with auto-detection of page type.

crawl_site

Trigger a fresh crawl to discover and analyze all site URLs and refresh data.