Steam Library MCP Server

A Model Context Protocol (MCP) server that provides access to your Steam game library data through Claude Desktop. It includes a helper script to copy your library data locally to a csv file for the MCP server to ingest.

This repo was developed with Claude Code, and I left Claude's config in here for reference. This was built simply as a learning experience and an example of how to create an MCP server.

Features

• Search Games: Find games by name, genre, developer, publisher, review summary, or maturity rating

• Filter Games: Filter by playtime, review summary, or maturity rating

• Game Details: Get comprehensive information about specific games

• Review Analysis: Detailed review statistics for games

• Library Statistics: Overview of your entire game library

• Recently Played: See what you've been playing lately

• Recommendations: Get game suggestions based on your playtime patterns

Related MCP server: steam-mcp

Example Interactions using Claude Desktop (Click the dropdowns to see responses)

Prerequisites

• Python 3.8 or higher

• A Steam account with a public game library

• Steam API key (get one from https://steamcommunity.com/dev/apikey)

• Your Steam ID

Setup

1. Install Dependencies

pip install -r requirements.txt

2. Fetch Your Steam Library Data

First, create a .env file with your Steam credentials:

# .env
STEAM_ID=your_steam_id_here
STEAM_API_KEY=your_steam_api_key_here

Then run the data fetcher:

python steam_library_fetcher.py

This will create a steam_library.csv file with all your game data.

3. Configure Claude Desktop

Copy the example configuration file and update the paths:

cp claude_desktop_config.example.json claude_desktop_config.json

Edit claude_desktop_config.json and update the paths to match your system:

{
  "mcpServers": {
    "Steam Library": {
      "command": "/path/to/your/python",
      "args": ["/path/to/your/simple-steam-mcp/mcp_server.py"],
      "env": {}
    }
  }
}

Then copy it to Claude Desktop's configuration location:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Linux: ~/.config/claude/claude_desktop_config.json Windows: %APPDATA%\claude\claude_desktop_config.json

4. Test the Server

You can test the server directly:

python mcp_server.py

5. Restart Claude Desktop

After updating the configuration file, restart Claude Desktop to load the MCP server.

Usage Examples

Once configured, you can ask Claude Desktop questions like:

• "What are my top 10 most played games?"

• "Show me all my puzzle games"

• "Find games with 'Very Positive' reviews that I haven't played yet"

• "What are some good games I should try based on what I've played?"

• "Show me details for Half-Life 2"

• "What games have I played recently?"

• "Give me statistics about my Steam library"

Available Tools

1. search_games: Search by name, genre, developer, publisher, review summary, or maturity rating

2. filter_games: Filter by playtime thresholds, review summary, or maturity rating

3. get_game_details: Get comprehensive info about a specific game

4. get_game_reviews: Get detailed review statistics

5. get_library_stats: Overview statistics of your library

6. get_recently_played: Games played in the last 2 weeks

7. get_recommendations: Personalized suggestions based on your playtime

Data Source

The server reads from steam_library.csv which should contain columns:

• appid, name, maturity_rating, review_summary, review_score, total_reviews

• positive_reviews, negative_reviews, genres, categories, developers, publishers

• release_date, playtime_forever, playtime_2weeks, rtime_last_played

Troubleshooting

1. Server not connecting: Check that the path in your Claude Desktop config is correct

2. CSV not found: Ensure steam_library.csv is in the same directory as the server script

3. Permission errors: Make sure Python has read access to the CSV file

4. Port conflicts: The server uses port 8000 by default - ensure it's available

Technical Details

• Built using the official MCP Python SDK

• Uses FastAPI for web transport with SSE (Server-Sent Events)

• Pandas for efficient CSV data processing

• Runs on all network interfaces (0.0.0.0) for flexibility