balldontlie

This MCP server integrates the Balldontlie API to provide data about players, teams, and games across NBA, NFL, and MLB. It exposes a set of tools you can query through your preferred LLMs or automation workflows to retrieve teams, players, games, and specific game details. The available tools are get_teams, get_players, get_games, and get_game. For example, you can fetch all NBA teams, search for players by name, or obtain game schedules for a given date range. The server can be invoked via npx balldontlie-mcp (often with an API key provided as an environment variable) and returns structured results you can feed into your prompts or applications. You can also use prompts like schedule_generator to generate interactive schedules for Claude Desktop. Overall, it enables building conversational or analytic flows that reference current data from the Balldontlie API within your Claude or other LLM-based assistants.

Prerequisites:

• Node.js and npm installed on your machine to use npx with the MCP package.
• An Balldontlie API key (free tier acceptable).

Installation steps:

1. Ensure Node.js is installed. Verify with: node -v npm -v

2. Install and run the MCP server via npx (no local installation required):

npx -y balldontlie-mcp

3. If you need to specify the API key via environment variable, set it in your shell before launching or include it in the mcp_config as shown in the example:

export BALLDONTLIE_API_KEY=your-api-key-here
npx -y balldontlie-mcp

4. Alternatively, install via Smithery as described in the README to integrate with Claude Desktop:

npx -y @smithery/cli install balldontlie-mcp --client claude

5. For Claude Desktop usage, configure claude_desktop_config.json as shown in the README (balldontlie with command npx and args [-y, balldontlie-mcp]).

Tips and caveats:

• Ensure the BALLDONTLIE_API_KEY (note the potential misspelling in docs) or BALLOY key is set as an environment variable where required. The example uses BALLDONTLIE_API_KEY in the config; double-check which env var name your deployment expects.
• The MCP exposes four main tools: get_teams, get_players, get_games, and get_game. Use the inputs described in the README to filter results by league (NBA, NFL, MLB) and to paginate with cursor where applicable.
• If you run into rate limits from Balldontlie, consider caching results or batching requests where possible.
• The schedule_generator prompt can help you generate an interactive schedule for Claude Desktop via Claude’s integration; keep in mind this requires Claude Desktop context and proper prompt formatting.
• This MCP server is described as MIT licensed; you may modify and redistribute per the license terms.
• If you need to switch to a different deployment (e.g., Docker, Python uv, etc.), you can adapt the mcp_config accordingly but this README uses the Node/npx path by default.