Game-predictor

You can run a Python-based MCP server that predicts video game scores from attributes like price, release timing, tags, genres, and platforms. It leverages machine learning to provide score predictions on a 0–100 scale and integrates data from Steam, RAWG, and SteamSpy to improve accuracy. This guide shows you how to use, install, and configure the server for practical MCP client interactions.

How to use

You interact with the server through an MCP client to request a predicted game score based on given characteristics. Provide inputs such as price, release month, tags, genres, and platforms, and the server returns a numeric score between 0 and 100. This enables quick estimates of how a game might be rated based on its attributes.

How to install

Prerequisites: ensure you have Python installed on your system (Python 3.8 or newer is recommended). You also need pip to install dependencies.

# 1) Ensure Python is installed
python --version

# 2) Install project dependencies
pip install -r requirements.txt

# 3) Set up environment variable for RAWG platform data
# Create a .env file or export the variable in your shell
# Example (use your actual API key):
RAWG_KEY=your_rawg_api_key_here

# 4) Run the MCP server
python mcp_server/main.py

Configuration and usage notes

To run the server locally, you will typically start the Python process pointing at the main script. The standard runtime command shown for configuring an MCP client is to execute Python with the path to the main server script and supply the RAWG API key via environment variables.

Claude Desktop configuration for MCP

{
  "mcpServers": {
    "game-predictor": {
      "command": "python",
      "args": ["/path/to/your/mcp_server/main.py"],
      "env": {
        "RAWG_KEY": "your_rawg_api_key_here"
      }
    }
  }
}

Troubleshooting and notes

If you encounter a missing dataset on first run, the server automatically creates the dataset from SteamSpy, Steam Store, and RAWG data. Ensure you have a valid RAWG API key set in your environment. If you see rate-limiting messages, the server includes built-in delays to respect API limits; wait and retry.

Available tools

make_prediction

Predicts a game's score based on its price, release month, tags, genres, and platforms using a machine learning model.