SportIntel

SportIntel MCP Server enables Claude and AI agents to access real-time, explainable sports analytics through a cohesive MCP interface. It provides player projections, lineup optimization, live odds, and SHAP-based explanations, all designed for seamless interaction with AI-powered decision-making workflows.

How to use

You will interact with SportIntel MCP Server through your MCP client by calling its tools to get projections, optimize lineups, retrieve live odds, or explain recommendations. Use each tool by passing the required sport and context, then consume the structured results to drive DFS decisions and analyses inside your AI workflows.

How to install

# Clone repository
git clone https://github.com/roizenlabs/sportintel-mcp.git
cd sportintel-mcp

# Install dependencies
npm install

# Set up environment
cp .env.example .env
# Edit .env with your API keys

Configuration for Claude Desktop

{
  "mcpServers": {
    "sportintel": {
      "command": "node",
      "args": ["/path/to/sportintel-mcp/dist/main.js"],
      "env": {
        "ODDS_API_KEY": "your_api_key_here"
      }
    }
  }
}

Run Standalone

# Development mode
npm run dev

# Production build
npm run build
npm start

Notes on usage and workflow

The SportIntel MCP Server exposes a compact set of tools you can invoke from your MCP client. Retrieve up-to-date projections with explainability, generate cash or GPP lineups under salary constraints, compare real-time odds across sportsbooks, and view explanations for each projection to understand the model’s reasoning.

Security and environment considerations

Ensure your API keys and credentials are stored securely in the environment file. Limit access to the Claude Desktop instance and rotate keys periodically. When deploying, prefer secure secret management practices and avoid hard-coding keys in public configurations.

Troubleshooting tips

If you encounter issues starting the server, verify that Node.js 18+ is installed, the dependencies are installed, and the path to the built main.js is correct in your Claude Desktop config. Check that ODDS_API_KEY is set in your environment and accessible to the runtime. Consult the console output for any startup errors and ensure the production build has been created before running npm start.

Available tools

get_player_projections

Fetch AI-powered DFS projections with SHAP explainability for a given slate and sport.

optimize_lineup

Perform multi-objective lineup optimization to generate optimal cash or GPP lineups under the salary cap.

get_live_odds

Retrieve real-time odds from multiple sportsbooks and compare lines across markets.

explain_recommendation

Provide SHAP-based explanations for projections and recommendations to understand feature impact.