Local SEO Data MCP Server

Model Context Protocol server for Local SEO Data. Gives AI assistants access to 42 local SEO tools -- SERP tracking, review monitoring, keyword research, AI visibility scoring, geogrid scans, and more.

One-time setup, then ask questions in plain English and your assistant calls the right endpoints automatically.

Prerequisites

1. Sign up at localseodata.com/signup
2. Create an API key in your API Keys dashboard
3. Replace YOUR_API_KEY below with your key (starts with sk_live_)

Setup

Claude Code

claude mcp add localseodata --transport http "https://mcp.localseodata.com/mcp?key=YOUR_API_KEY" --scope user

The --scope user flag makes it available across all your projects. Restart Claude Code after adding.

Claude Desktop

Add to your config file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "localseodata": {
      "url": "https://mcp.localseodata.com/mcp?key=YOUR_API_KEY"
    }
  }
}

Restart Claude Desktop after saving.

Cursor

Add to .cursor/mcp.json in your project root, or ~/.cursor/mcp.json for global access:

{
  "mcpServers": {
    "localseodata": {
      "url": "https://mcp.localseodata.com/mcp?key=YOUR_API_KEY"
    }
  }
}

VS Code (GitHub Copilot)

Add to .vscode/mcp.json in your project:

{
  "servers": {
    "localseodata": {
      "type": "streamable-http",
      "url": "https://mcp.localseodata.com/mcp?key=YOUR_API_KEY"
    }
  }
}

Cline

Open Cline settings > MCP Servers > Add Remote Server:

• Name: localseodata
• URL: https://mcp.localseodata.com/mcp?key=YOUR_API_KEY

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "localseodata": {
      "serverUrl": "https://mcp.localseodata.com/mcp?key=YOUR_API_KEY"
    }
  }
}

Available Tools

| Category | Tools | |----------|-------| | SERP Data | local_pack, organic_serp, local_finder, maps, local_services_ads | | AI/LLM Optimization | ai_overview, ai_mode, ai_mentions, ai_visibility, ai_compare, ai_keyword_data, ai_top_sources, ai_top_pages | | Business Data | business_profile, google_reviews, qa, business_listings, multi_platform_reviews | | Keyword Research | search_volume, keyword_suggestions, keywords_for_site, related_keywords, keyword_trends | | Audits | local_audit, citation_audit, reputation_audit, page_audit, profile_health | | Competitive Intel | competitor_gap, competitor_ads, brand_mentions, backlink_gap, backlink_summary | | Reports | review_velocity, local_authority, keyword_opportunities | | Geogrid | geogrid_scan | | Locations | location_search | | Utility | ping, get_balance |

Example Prompts

Once connected, just ask naturally:

• "Who ranks in the local pack for 'plumber' in Denver?"
• "Run a full local SEO audit for Mike's Plumbing in Austin, TX"
• "Compare my Google reviews to my top 3 competitors"
• "What keywords should a dentist in Portland be targeting?"
• "Check if my business NAP is consistent across directories"
• "Map my rankings on a 5x5 geogrid for 'bakery' in Brooklyn"
• "What's my AI visibility score for 'best pizza in Chicago'?"

Authentication

Your API key is passed as a URL query parameter (?key=YOUR_API_KEY). This works with every MCP client without needing custom header configuration. Keys can be created or revoked anytime in the API Keys dashboard.

Pricing

MCP calls consume credits from your Local SEO Data account, same as REST API calls. Most endpoints cost 1 credit ($0.005). Monitor usage in your dashboard.

See localseodata.com/pricing for plans.

Self-Hosting

If you want to run the MCP server yourself:

git clone https://github.com/localseodata/mcp-server.git
cd mcp-server
npm install
npm run build

Set the required environment variable and start:

export API_BASE_URL=https://api.localseodata.com
npm start

The server runs on port 3003 by default (set PORT to change). A Dockerfile is included for container deployments.

Development

npm run dev       # Watch mode with hot reload
npm run build     # Compile TypeScript
npm run typecheck # Type check without emitting
npm start         # Run compiled output

License

MIT