league-analysis-mcp-server 0.2.2

Model Context Protocol server for Yahoo Fantasy Sports API with advanced historical analysis and manager profiling

League Analysis MCP Server

A comprehensive Model Context Protocol (MCP) server that provides AI assistants with access to Yahoo Fantasy Sports data, including advanced historical analysis and manager profiling capabilities.

Features

🏈 Multi-Sport Support

• NFL (National Football League)
• NBA (National Basketball Association)
• MLB (Major League Baseball)
• NHL (National Hockey League)

📊 Current Season Data

• League information and settings
• Real-time standings
• Team rosters and lineups
• Weekly matchups and scoring
• Transaction history

📈 Historical Analysis

• Multi-season draft analysis - Track draft patterns and strategies over time
• Manager performance history - Comprehensive performance metrics across seasons
• Transaction pattern analysis - Trading behavior and partnership identification
• Season-to-season comparisons - League evolution and competitive balance trends

🧠 Advanced Analytics

• Draft strategy classification - Identify RB-heavy, Zero-RB, or balanced approaches
• Manager skill evaluation - Comprehensive skill scoring based on multiple metrics
• Trade likelihood prediction - Predict trade partnerships based on historical patterns
• Pattern recognition - Identify trends in manager behavior and league dynamics

⚡ Performance Features

• Smart caching - Historical data cached permanently, current data with TTL
• Rate limiting - Respects Yahoo API limits
• Error handling - Comprehensive error handling and logging
• Multi-season support - Access data from 2015+ seasons

Installation

Prerequisites

• Python 3.10+
• uv installed (Python package manager)
• Yahoo Developer App credentials (setup will guide you)

🚀 Install uv (Required)

# Install uv (one-time setup)
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows: powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

✨ No Package Installation Required!

The server automatically downloads when you configure your MCP client. Just add the configuration below - no separate package installation needed!

How it works:

1. You install uv (the package manager)
2. You configure your MCP client to use uvx league-analysis-mcp-server
3. When your MCP client starts, uvx automatically downloads the server from PyPI
4. Subsequent starts use the cached version

🔧 Development Installation

From Source:

git clone https://github.com/ari1110/League-Analysis-MCP.git
cd League-Analysis-MCP
uv sync --all-extras

Then set up authentication using the conversational MCP tools (see streamlined setup below).

🚀 NEW! Streamlined Authentication Setup

No more complex setup scripts! Authentication is now handled entirely through conversational MCP tools with two OAuth options:

Option 1: Automated OAuth (Recommended)

5-Step Setup Process:

1. Install & Connect - Add server to your MCP client
2. Check Status - AI assistant runs check_setup_status()
3. Create Yahoo App - AI assistant shows you exactly what to do with create_yahoo_app()
4. Save Credentials - AI assistant saves them for you with save_yahoo_credentials(key, secret)
5. Automated OAuth - AI assistant runs start_automated_oauth_flow() for fully automated setup

✨ Fully Automated Experience:

• Browser opens automatically to Yahoo authorization page
• You sign in and authorize the app
• HTTPS callback server automatically captures the authorization code
• Tokens are saved and exchanged automatically
• Success page displays and auto-closes
• Total time: ~30 seconds!

Option 2: Manual OAuth (Fallback)

5-Step Setup Process:

1. Install & Connect - Add server to your MCP client
2. Check Status - AI assistant runs check_setup_status()
3. Create Yahoo App - AI assistant shows you exactly what to do with create_yahoo_app()
4. Save Credentials - AI assistant saves them for you with save_yahoo_credentials(key, secret)
5. Manual OAuth - AI assistant guides you through start_oauth_flow() and complete_oauth_flow(code)

📋 Manual Code Entry:

• AI provides authorization URL
• You visit URL, sign in, and get verification code
• You provide the code to the AI assistant
• AI completes token exchange

That's it! ✨ Everything happens in your conversation with the AI assistant. No file editing, no command line scripts, no leaving the interface.

🔑 Important OAuth Requirements

For Best Results:

• ✅ Sign into Yahoo first: Before starting OAuth, sign into yahoo.com in your browser
• ✅ Use the same account: Sign in with the Yahoo account that has your fantasy leagues
• ✅ Stay signed in: Keep your Yahoo session active during the OAuth process

Why this matters:

• Automated OAuth works best when you're already signed into Yahoo
• Fresh logins during OAuth can sometimes cause Yahoo's "uh-oh" errors
• Existing sessions provide the smoothest authorization experience

SSL Certificate Handling: When using automated OAuth, your browser may show a security warning for https://localhost:8080. This is normal and safe:

1. Click "Advanced" or "Show Details"
2. Click "Proceed to localhost (unsafe)" or similar
3. The success page will display and auto-close after 3 seconds

Troubleshooting:

• "uh-oh" errors: Try signing into Yahoo first, then retry the OAuth flow
• SSL warnings: These are normal for self-signed certificates - safe to proceed
• Timeouts: Use manual OAuth flow as fallback if automated flow has issues

Example User Experience:

You: "Show me my fantasy league standings"

AI: Let me check that for you. First, I need to check your authentication setup.
    → Runs check_setup_status()
    
AI: I see you need to create a Yahoo Developer app first. Here are the exact steps:
    → Runs create_yahoo_app()
    → Shows step-by-step instructions with exact values to use

You: "I created the app! My key is dj0yJmk9... and secret is abc123..."

AI: Perfect! Let me save those credentials for you.
    → Runs save_yahoo_credentials(key, secret)
    → Automatically saves to your environment

AI: Great! Now let's complete the OAuth authorization. Please visit this URL:
    → Runs start_oauth_flow()
    → Shows authorization URL and clear instructions

You: "I authorized it and got code: xyz789"

AI: Excellent! Let me complete the setup.
    → Runs complete_oauth_flow("xyz789")
    
AI: 🎉 Setup complete! Now let me get your league standings...
    → Runs get_standings() and shows your data

Total time: ~2 minutes. No technical knowledge required!

🔧 Development Setup (Advanced Users)

For development or troubleshooting, you can also run the server manually after installing dependencies with uv sync --all-extras.

Usage

Starting the Server

# PyPI Installation (recommended):
uvx league-analysis-mcp-server

# Or if installed with pip:
league-analysis-mcp-server

# Development/Source:
uv run python -m src.server

🔌 MCP Client Configuration

For detailed setup instructions for all MCP clients, see MCP_INTEGRATION_GUIDE.md

Quick setup for Claude Desktop: Add to claude_desktop_config.json:

{
  "mcpServers": {
    "league-analysis": {
      "command": "uvx",
      "args": ["league-analysis-mcp-server"]
    }
  }
}

The integration guide covers configuration for Claude Desktop, Claude Code, Continue.dev, and other MCP clients with complete examples and troubleshooting.

🔧 Testing Your Connection

After adding to your MCP client:

1. Restart your MCP client (Claude Desktop, etc.)
2. Test the connection by asking:
   • "Can you get server info for the league analysis server?"
   • "List available seasons for NFL"
   • "What fantasy sports tools are available?"
3. Check server status with: get_server_info()

🚨 Troubleshooting MCP Connection

Common issues:

1. Server not found:

   • Check the file path in your config
   • Ensure UV is installed and in PATH
   • Verify the server starts manually: uv run python -m src.server

2. Authentication errors:

   • Use the streamlined setup: Ask your AI assistant to run check_setup_status()
   • Follow the conversational setup process using MCP tools
   • Check your Yahoo Developer app settings (redirect URI must be https://localhost:8080/)

3. Permission issues:

   • Ensure your MCP client has permission to execute UV
   • Check working directory permissions

4. Environment variables:

   • MCP clients may not inherit your shell environment
   • Create .env file in project root with credentials

Available Tools

🔐 Streamlined Authentication Tools ✨ NEW!

• check_setup_status() - Check current authentication state and get next steps
• create_yahoo_app() - Step-by-step Yahoo Developer app creation guide
• save_yahoo_credentials(consumer_key, consumer_secret) - Save Yahoo app credentials
• start_automated_oauth_flow() - 🚀 Fully automated OAuth with callback server (recommended)
• start_oauth_flow() - Begin manual OAuth authorization with clear instructions
• complete_oauth_flow(verification_code) - Complete manual setup with verification code
• test_yahoo_connection() - Test API connectivity and troubleshoot issues
• reset_authentication() - Clear all auth data to start fresh

Basic League Tools

• get_server_info() - Server status and configuration
• get_setup_instructions() - Comprehensive setup help (includes new tools guidance)
• list_available_seasons(sport) - Available historical seasons
• get_league_info(league_id, sport, season?) - League settings and metadata
• get_standings(league_id, sport, season?) - Current or historical standings
• get_team_roster(league_id, team_id, sport, season?) - Team roster information
• get_matchups(league_id, sport, week?, season?) - Weekly matchup data

Historical Analysis Tools

• get_historical_drafts(league_id, sport, seasons?) - Draft results across seasons
• get_season_transactions(league_id, sport, season) - Transaction history for season
• analyze_manager_history(league_id, sport, seasons?, team_id?) - Manager performance patterns
• compare_seasons(league_id, sport, seasons) - Season-to-season analysis

Advanced Analytics Tools

• analyze_draft_strategy(league_id, sport, seasons?, team_id?) - Draft pattern analysis
• predict_trade_likelihood(league_id, sport, team1_id?, team2_id?, seasons?) - Trade predictions
• evaluate_manager_skill(league_id, sport, seasons?, team_id?) - Comprehensive skill evaluation

Cache Management

• clear_cache(cache_type?) - Clear cached data ('all', 'current', 'historical')

Available Resources

Access read-only data through these resource URIs:

• league_overview://sport/league_id[/season] - Comprehensive league overview
• current_week://sport/league_id - Current week activity and focus areas
• league_history://sport/league_id - Multi-season history and trends
• manager_profiles://sport/league_id[/team_id] - Manager profiling information

Example Usage

# Get basic league info
result = get_league_info("123456", "nfl")

# Analyze manager performance across last 3 seasons  
analysis = analyze_manager_history("123456", "nfl", ["2022", "2023", "2024"])

# Get draft strategies for all managers
draft_analysis = analyze_draft_strategy("123456", "nfl", ["2022", "2023", "2024"])

# Evaluate manager skill levels
skill_eval = evaluate_manager_skill("123456", "nfl", ["2022", "2023", "2024"])

# Predict trade likelihood between specific managers
trade_pred = predict_trade_likelihood("123456", "nfl", "team1", "team2")

Configuration

Game IDs

The server includes game ID mappings for seasons 2015-2024 across all supported sports. These are automatically used when specifying historical seasons.

Caching Strategy

• Historical data: Cached permanently (TTL = -1)
• Current season data: Cached for 5 minutes (TTL = 300)
• Cache size: Limited to 100MB by default

Rate Limiting

• Default: 60 requests per minute
• Burst limit: 10 requests
• Automatically handled by the server

Architecture

Core Components

• FastMCP 2.0: High-level MCP framework for rapid development
• YFPY: Yahoo Fantasy Sports API wrapper
• Caching Layer: Smart caching for performance optimization
• Authentication Manager: OAuth handling for Yahoo API access
• Analytics Engine: Advanced pattern recognition and predictions

Data Flow

1. Request → Authentication → Cache Check → Yahoo API → Response Processing → Cache Storage → Response
2. Historical Data: Cached permanently after first fetch
3. Current Data: Cached with TTL, automatically refreshed

Error Handling

• Comprehensive error logging
• Graceful degradation for missing data
• Cache fallback for API failures
• User-friendly error messages

Supported Analysis Types

Manager Profiling

• Performance Tiers: Elite, Above Average, Average, Below Average, Needs Improvement
• Consistency Scoring: Win rate, scoring, and ranking consistency
• Success Patterns: Championship rate, playoff appearances, trajectory analysis

Draft Strategy Classification

• RB-Heavy: Prioritizes running backs in early rounds
• Zero-RB: Waits on running backs, focuses on WR/other positions
• Balanced: Even distribution across position types
• Auction Analysis: Spending patterns and value identification

Trade Pattern Analysis

• Partnership Identification: Historical trade frequency between managers
• Likelihood Scoring: Probability of future trades based on history
• Trade Timing: Seasonal patterns and deadline behavior

Troubleshooting

Common Issues

1. Authentication Errors

   • ✅ NEW: Use streamlined setup by asking AI to run check_setup_status()
   • 🚀 Recommended: Try start_automated_oauth_flow() for easiest setup
   • ⚠️ "uh-oh" errors: Sign into yahoo.com first, then retry OAuth
   • 🔄 Reset & retry: Use reset_authentication() to start fresh
   • 🔧 Manual setup: Verify Yahoo Consumer Key/Secret in .env
   • 📝 Help: Run get_setup_instructions() for detailed guidance

2. OAuth Troubleshooting

   • SSL certificate warnings: Normal for automated OAuth - safe to proceed
   • Timeout during OAuth: Try manual OAuth as fallback (start_oauth_flow())
   • Yahoo session issues: Sign into Yahoo first, keep session active
   • Port conflicts: Ensure nothing else uses port 8080
   • Browser compatibility: Modern browsers work best (Chrome, Edge, Firefox)

3. No Historical Data

   • Ensure league has existed for multiple seasons
   • Verify correct league ID and sport combination
   • Check game ID mappings in config/game_ids.json

4. Cache Issues

   • Use clear_cache("all") to reset all cached data
   • Check cache statistics with get_server_info()

5. Rate Limiting

   • Server automatically handles rate limits
   • Historical data queries may take longer due to multiple API calls
   • Use caching to minimize repeated requests

Support

For issues and feature requests, please check the documentation or create an issue in the project repository.

License

MIT License - see LICENSE file for details.

Contributing

Contributions welcome! Please read the contributing guidelines and submit pull requests for any improvements.