Model Context Protocol (MCP)

Give Claude, Cursor, and other AI assistants real-time access to your Google search rankings, project stats, and keyword metrics via our official MCP server.

Installation

The @trackserps/mcp package runs locally over stdio and communicates with the TrackSERPs REST API using your API key.

Terminal

npm install -g @trackserps/mcp

Configuration

1. Claude Desktop

Add the trackserps server to your Claude Desktop configuration file (found at ~/Library/Application Support/Claude/claude_desktop_config.json on macOS or %APPDATA%\\Claude\\claude_desktop_config.json on Windows):

claude_desktop_config.json

{
  "mcpServers": {
    "trackserps": {
      "command": "npx",
      "args": ["-y", "@trackserps/mcp"],
      "env": {
        "TRACKSERPS_API_KEY": "ts_live_your_api_key_here"
      }
    }
  }
}

2. Claude Code CLI

Configure Claude Code with a single CLI command:

Claude Code Command

claude mcp add trackserps -- npx -y @trackserps/mcp

3. Cursor & Windsurf

In Cursor Settings → Features → MCP, add a new server with type command, command npx -y @trackserps/mcp, and environment variable TRACKSERPS_API_KEY set to your token.

Available AI Tools

Once configured, your AI assistant can execute the following tools autonomously:

Tool Name	Description	Action Type
`list_projects`	Lists all projects tracked by your account with keyword counts and domains.	Read
`find_location`	Searches geographic location codes and country codes for targeted SERP scans.	Read
`get_rankings`	Fetches latest rankings across desktop and mobile for a project.	Read
`get_ranking_movement`	Calculates top gainers, drops, and net average movement across a date range.	Read
`get_keyword_history`	Fetches daily ranking timeline and delta change for a single keyword.	Read
`get_search_volume`	Retrieves historical monthly search volume, estimated CPC, and competition.	Read
`get_usage`	Inspects monthly check credits remaining and API quota consumption.	Read
`add_keywords`	Upserts keywords to a project. Returns cost estimate unless `confirm: true` is passed.	Write (Dry Run safe)
`run_check`	Triggers a manual ranking check run. Returns credit estimate unless `confirm: true` is passed.	Write (Dry Run safe)
`get_run_status`	Monitors execution status of an active check run.	Read

Built-in Safety Safeguards

Dry-Run by Default: All write operations (such as add_keywords and run_check) require explicit confirm: true to execute. If omitted, the tool performs a safe dry-run calculation, outputting the required credits without deducting from your account balance.