GameAnalytics MCP Server

The GameAnalytics Model Context Protocol (MCP) server enables Large Language Models (LLMs) to natively access your game's data through the GameAnalytics Metrics API. This allows you to ask questions about your game's performance, user retention, and revenue in natural language.

---

Overview

This integration wraps the GameAnalytics Metrics API, exposing 33 specialized tools that allow AI assistants to:

• Discover Data: Explore available games, studios, dimensions, and metrics.
• Analyze Performance: Retrieve key KPIs like Retention, DAU/MAU, and Revenue.
• Debug & Optimize: Investigate ad impressions, session lengths, and conversion rates.

All data is fetched in real-time using your secure GameAnalytics Metrics API key.

---

MCP Tools

The server exposes the following tools, categorized by function:

Data Discovery & Exploration

These tools help the AI understand your data structure before querying specific metrics.

• get_games: List available games for a time period.
• get_studios: List available studios.
• get_dimensions: List supported dimensions (e.g., country_code, platform).
• get_dimension_values: Get valid values for a specific dimension.
• get_interval: Check data availability dates.
• get_supported_metrics: List all queryable metrics.

Revenue Analysis

• fetch_revenue: Total revenue from in-app purchases.
• fetch_arpu: Average Revenue Per User (includes ARPDAU, ARPWAU, ARPMAU).
• fetch_arppu: Average Revenue Per Paying User.
• fetch_paying_users: Count of Paying Users.
• fetch_converting_users: Count of Users performing their first payment event.
• fetch_conversion: The percent of users who made a purchase for the specified time interval.
• fetch_conversion_ft: The percent of users who made their first purchase on the specified time interval.

Impression Level Revenue Data Analysis

• fetch_ilrd_revenue: Impression Level Revenue Data (Ad Revenue).
• fetch_ilrd_arpu: revenue sum per user reported by ad network providers.
• fetch_ilrd_revenue_per_session: revenue per session from ad network providers.
• fetch_ilrd_revenue_per_active_user: revenue per daily active user from ad network providers.
• fetch_ilrd_count_per_active_user: revenue per daily active user from ad network providers.

User Engagement

• fetch_user_unique: Active users (DAU, WAU, MAU).
• fetch_new_users: Count of new installs.
• fetch_installs_ratio: Ratio of new installs.
• fetch_retention_ahead: Standard retention (D1, D7, D30, etc.).
• fetch_retention_retro: The percent of users who showed up again X days after installing the game.
• fetch_retention_count: The count of users who showed up again X days after installing the game.
• fetch_returning_users: Count of users returning after X days.
• fetch_session_length: Average session duration.
• fetch_sessions_per_user: Frequency of play sessions.
• fetch_session_unique: The count of sessions played on the specified time interval.
• fetch_session_length_per_user: The amount of time spent within the app.
• fetch_event_count: Count of events.

Ad Monetization

• fetch_ad_impressions_count: Total ads shown.
• fetch_ad_impressions_per_session: Ad density per session.
• fetch_ad_impressions_per_user: Ad density per user.

---

Getting Started

Prerequisites

• Python: Version 3.11 or higher.
• API Key: A valid GameAnalytics Metrics API Key. You can generate it in your GameAnalytics Metrics API Settings (PipelineIQ → Metrics API → Generate Key).

Installation

1. Install with pip:

   pip install git+https://github.com/GameAnalytics/GA-MCP

2. Set your API Key:

   export GameAnalytics_API_KEY="your_api_key_here"

---

Implementation Guide

Option 1: Claude Desktop

To use this server with Claude Desktop, add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "GameAnalytics": {
      "command": "python3",
      "args": ["-m", "ga_mcp.server"],
      "env": {
        "GAMEANALYTICS_API_KEY": "your_api_key_here"
      }
    }
  }
}

Option 2: Cursor

1. Go to Cursor Settings > Features > MCP Servers.
2. Click + Add New MCP Server.
3. Name: GameAnalytics
4. Type: command
5. Command: python3 -m ga_mcp.server
6. Environment Variables: Add GAMEANALYTICS_API_KEY with your key.

---

Use Case Examples

Once connected, you can ask your AI assistant questions like:

Basic Performance

"What was the DAU and Revenue for 'My Game' last week?"

Retention Analysis

"Compare D1 and D7 retention for iOS vs Android users for the last month."

Monetization Deep Dive

"Show me the ARPDAU trend for the last 30 days, split by country."

Ad Strategy

"Is there a correlation between session length and ad impressions per user?"

---

Security & Compliance

• Read-Only Access: This MCP server is designed for read-only access to your metrics. It cannot modify your game data or settings.
• Local Execution: The server runs locally on your machine. Your API key is stored in your local configuration and is never sent to Anthropic, OpenAI, or Google.
• Data Privacy: Data retrieved from GameAnalytics is passed directly to the LLM context for analysis. Ensure you comply with your organization's data handling policies when using LLMs.