sensortower-mcp 1.2.10

Model Context Protocol server for Sensor Tower APIs using FastMCP

Sensor Tower MCP Server

Access Sensor Tower's mobile app intelligence APIs through the Model Context Protocol (MCP). Get app store data, rankings, downloads, revenue estimates, and competitor intelligence directly in MCP-compatible tools like Cursor.

Features

• 📱 App Intelligence - Metadata, rankings, downloads, revenue estimates
• 🔍 Search & Discovery - Find apps and publishers by name or description
• 📊 Market Analysis - Category rankings, featured apps, competitor insights
• 🛠️ Developer Tools - Access your own app analytics and metrics

Quick Start

# Install with uvx (recommended)
uvx sensortower-mcp

# Or install with pip
pip install sensortower-mcp

MCP Configuration

Add to your MCP settings (e.g., Cursor):

{
  "mcpServers": {
    "sensortower": {
      "command": "uvx",
      "args": ["sensortower-mcp"],
      "env": {
        "SENSOR_TOWER_API_TOKEN": "your_api_token_here"
      }
    }
  }
}

API Token

Get your API token from Sensor Tower Account Settings.

Available Tools

Tool	Description
get_app_metadata	App details, ratings, categories
search_entities	Search apps and publishers
get_category_rankings	Top apps by category
get_download_estimates	Download trends and estimates
get_revenue_estimates	Revenue data and forecasts
get_creatives	Advertising creatives (requires ad_types)
get_country_codes	Available country codes
get_category_ids	Platform category IDs

Example Usage

# Search for social media apps
search_entities(os="ios", entity_type="app", term="social media", limit=10)

# Get advertising creatives (ad_types is required by Sensor Tower)
get_creatives(
    os="ios",
    app_ids="284882215",
    start_date="2024-01-01",
    countries="US",
    networks="Instagram",
    ad_types="video"
)

# Get top free apps in Social category
get_category_rankings(
    os="ios",
    category="6005",
    chart_type="topfreeapplications",
    country="US",
    date="2024-01-15"
)

Requirements

• Python 3.10+
• Sensor Tower API token
• MCP-compatible client (Cursor, etc.)

Docker

# Run with Docker
docker run -e SENSOR_TOWER_API_TOKEN="your_token" -p 8666:8666 \
  bobbysayers492/sensortower-mcp sensortower-mcp --transport http --port 8666

Links

• Docker Hub
• Full Documentation
• Sensor Tower API
• Model Context Protocol

License

MIT License