MLB API MCP Server

A Model Context Protocol (MCP) server that provides comprehensive access to MLB statistics and baseball data through a FastMCP-based interface.

Overview

This MCP server acts as a bridge between AI applications and MLB data sources, enabling seamless integration of baseball statistics, game information, player data, and more into AI workflows and applications.

Features

MLB Data Access

• Current standings for all MLB teams with flexible filtering by league, season, and date
• Game schedules and results with date range support
• Player statistics including traditional and sabermetric stats (WAR, wOBA, wRC+)
• Team information and rosters with various roster types
• Live game data including boxscores, linescores, and play-by-play
• Game highlights and scoring plays
• Player and team search functionality
• Draft information and award recipients
• Game pace statistics and lineup information

MCP Tools

All MLB/statistics/game/player/team/etc. functionality is exposed as MCP tools, not as RESTful HTTP endpoints. These tools are accessible via the /mcp/ endpoint using the MCP protocol. For a list of available tools and their descriptions, visit /tools/ when the server is running.

Key MCP Tools

• get_mlb_standings - Current MLB standings with league and season filters
• get_mlb_schedule - Game schedules for specific dates, ranges, or teams
• get_mlb_team_info - Detailed team information
• get_mlb_player_info - Player biographical information
• get_mlb_boxscore - Complete game boxscores
• get_mlb_linescore - Inning-by-inning game scores
• get_mlb_game_highlights - Video highlights for games
• get_mlb_game_scoring_plays - Play-by-play data with event filtering
• get_mlb_game_pace - Game duration and pace statistics
• get_mlb_game_lineup - Detailed lineup information for games
• get_multiple_mlb_player_stats - Traditional player statistics
• get_mlb_sabermetrics - Advanced sabermetric statistics (WAR, wOBA, etc.)
• get_mlb_roster - Team rosters with various roster types
• get_mlb_search_players - Search players by name
• get_mlb_search_teams - Search teams by name
• get_mlb_players - All players for a sport/season
• get_mlb_teams - All teams for a sport/season
• get_mlb_draft - Draft information by year
• get_mlb_awards - Award recipients
• get_current_date - Current date
• get_current_time - Current time

For the full list and detailed descriptions, see /tools/ or /docs when the server is running.

HTTP Endpoints

The following HTTP endpoints are available:

• / - Redirects to /docs
• /docs - Interactive API documentation and tool listing
• /health/ - Health check endpoint
• /mcp/info - MCP server information
• /tools/ - List of all available MCP tools
• /mcp/ (POST) - MCP protocol endpoint for MCP-compatible clients

Note: There are no RESTful HTTP endpoints for MLB/statistics/game/player/team/etc. All such functionality is accessed via MCP tools through the /mcp/ endpoint.

MCP Integration

• Compatible with MCP-enabled AI applications
• Tool-based interaction model with comprehensive endpoint descriptions
• Automatic API documentation generation
• Schema validation and type safety
• Full response schema descriptions for better AI integration

Installation

Installing via Smithery

To install MLB API Server for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @guillochon/mlb-api-mcp --client claude

Option 1: Local Installation

1. Install uv if you haven't already:

curl -LsSf https://astral.sh/uv/install.sh | sh

2. Clone the repository:

git clone https://github.com/guillochon/mlb-api-mcp.git
cd mlb-api-mcp

3. Create and activate a virtual environment:

uv venv
source .venv/bin/activate  # On Unix/macOS
# or
.venv\Scripts\activate  # On Windows

4. Install dependencies:

uv pip install -e .

Option 2: Docker Installation

1. Clone the repository:

git clone https://github.com/guillochon/mlb-api-mcp.git
cd mlb-api-mcp

2. Build the Docker image:

docker build -t mlb-api-mcp .

3. Run the container (default timezone is UTC, uses Python 3.12):

docker run -p 8000:8000 mlb-api-mcp

Setting the Timezone

To run the container in your local timezone, pass the TZ environment variable (e.g., for New York):

docker run -e TZ=America/New_York -p 8000:8000 mlb-api-mcp

Replace America/New_York with your desired IANA timezone name.

The server will be available at http://localhost:8000 with:

• MCP Server: http://localhost:8000/mcp/
• Documentation: http://localhost:8000/docs

Docker Options

You can also run the container with additional options:

# Run in detached mode
docker run -d -p 8000:8000 --name mlb-api-server mlb-api-mcp

# Run with custom port mapping
docker run -p 3000:8000 mlb-api-mcp

# View logs
docker logs mlb-api-server

# Stop the container
docker stop mlb-api-server

# Remove the container
docker rm mlb-api-server

Usage

Starting the Server

Run the MCP server locally:

# For stdio transport (default, for MCP clients like Smithery)
uv run python main.py

# For HTTP transport (for web access)
uv run python main.py --http

The server will start with:

• MCP Server on http://localhost:8000/mcp/
• Interactive API documentation available at http://localhost:8000/docs

MCP Client Integration

This server can be integrated into any MCP-compatible application. The server provides tools for:

• Retrieving team standings and schedules
• Getting comprehensive player and team statistics
• Accessing live game data and historical records
• Searching for players and teams
• Fetching sabermetric statistics like WAR
• And much more...

API Documentation

Once the server is running, visit http://localhost:8000/docs for comprehensive API documentation including:

• Available HTTP endpoints
• List of all available MCP tools at /tools/
• Tool descriptions and parameters
• Interactive testing interface
• Parameter descriptions and examples

Dependencies

• mcp[cli]: MCP-compliant server framework with CLI support
• FastAPI: Web framework for HTTP transport
• python-mlb-statsapi: Official MLB Statistics API wrapper
• uvicorn[standard]: ASGI server for running the app
• websockets: WebSocket support (latest version to avoid deprecation warnings)
• python-dotenv: Environment variable management
• httpx: HTTP client for API requests

Development

This project uses:

• Python 3.10+ (Docker uses Python 3.12)
• FastMCP for the web framework
• uv for fast Python package management
• Hatchling for build management
• MLB Stats API for comprehensive baseball data access
• Ruff for linting and formatting

Setup Pre-commit Hooks

1. Install pre-commit:

pip install pre-commit

2. Initialize pre-commit hooks:

pre-commit install

Now, the linting checks will run automatically whenever you commit code. You can also run them manually:

pre-commit run --all-files

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

License

This project is open source. Please check the license file for details.

Testing

This project includes comprehensive test coverage with pytest and coverage reporting.

Running Tests

# Run all tests with coverage (default)
uv run pytest

# Run tests with verbose output
uv run pytest -v

# Run specific test file
uv run pytest tests/test_mlb_api.py

# Run specific test function
uv run pytest tests/test_mlb_api.py::test_get_mlb_standings

# Run tests without coverage
uv run tests/run_coverage.py test

# Generate HTML coverage report
uv run tests/run_coverage.py html

# Clean up coverage files
uv run tests/run_coverage.py clean

Coverage

• Current Coverage: 89.91% (exceeds 80% threshold)
• Coverage Source: mlb_api.py and generic_api.py
• Reports: Terminal output, HTML (htmlcov/index.html), and XML (coverage.xml)
• CI Integration: Coverage checking runs automatically on every push/PR; results are published in the GitHub Actions job summary

Test Structure

The test suite includes:

• Unit tests for all MCP tools (MLB API and Generic API)
• Error handling tests for API failures
• Edge case tests for boundary conditions
• Mock-based tests to avoid external API calls

Live-API Scan

The mocked test suite protects this repo's own logic but does not exercise the real upstream MLB Stats API / pybaseball services, whose response shapes are outside this repo's control. To verify the tools still work against live data (e.g. after upstream field renames or contract changes), run the one-time scan:

# Invokes every registered tool against the real upstream API and prints a
# pass/fail summary plus the top-level return type of each result.
uv run python scripts/scan_live_tools.py

The scan writes a full JSON report to scripts/live_scan_report.json (gitignored). This is a manual verification pass, not part of the automated CI suite.

Adding New Tests

When adding new functionality: