KicktippAi - Kicktipp.de Automation POC

A C# proof-of-concept for automating login and betting on kicktipp.de, inspired by the Python project schwalle/kicktipp-betbot and its fork ehonda/kicktipp-cli.

Features

✅ Completed:

• Environment-based credentials: Load username/password from .env file
• Secure login automation: Uses HttpClient and AngleSharp for form parsing and submission
• Cookie management: Extracts and saves login tokens for future use
• Open predictions fetching: Retrieves available matches for betting
• Random bet placement: Automatically places random bets on open predictions
• Safety features: Dry-run mode and existing bet detection
• Hard-coded community: Currently targets "ehonda-test" community

Architecture

• HttpClient + AngleSharp: Modern C# web automation stack
• Environment variables: Secure credential storage via .env file
• HTTPS support: All communication uses secure connections
• Form parsing: Robust HTML form handling for login and betting
• Error handling: Comprehensive exception handling and status reporting

Usage

Setup Credentials

For security, credentials are stored outside the repository to prevent AI agents with solution directory read access from accidentally leaking credentials to remote sources:

1. Create secrets directory: A KicktippAi.Secrets directory should exist as a sibling to the solution directory
2. Setup credentials: Copy .env.example to the secrets directory at KicktippAi.Secrets/dev/Poc/.env
3. Add your credentials: Edit the .env file with your actual kicktipp.de credentials

# Directory structure:
├── KicktippAi/                 # This repository
│   ├── dev/Poc/.env.example    # Template file
│   └── ...
└── KicktippAi.Secrets/         # Secrets directory (outside repo)
    └── dev/Poc/.env            # Your actual credentials

Running the Application

1. Navigate to project: cd dev/Poc
2. Run the application: dotnet run
3. Interactive betting: The app will show a dry-run first, then ask for confirmation

Example Output

Kicktipp.de Automation POC
==========================
✓ Login successful!
✓ Login token extracted and saved to .env file

✓ Found 4 open matches:
  04.07.2025 21:00 'Fluminense' vs. 'Al-Hilal'
  05.07.2025 03:00 'Palmeiras' vs. 'FC Chelsea'
  05.07.2025 18:00 'Paris St. Germain' vs. 'FC Bayern München'
  05.07.2025 22:00 'Real Madrid' vs. 'Borussia Dortmund'

=== DRY RUN ===
05.07.2025 03:00 'Palmeiras' vs. 'FC Chelsea' - betting 3:0
05.07.2025 18:00 'Paris St. Germain' vs. 'FC Bayern München' - betting 0:0
05.07.2025 22:00 'Real Madrid' vs. 'Borussia Dortmund' - betting 1:2
Summary: 3 bets to place, 0 skipped

Do you want to place these bets for real? (y/N): y

=== PLACING REAL BETS ===
✓ Successfully submitted 3 bets!

Technical Implementation

Random Bet Generation

The SimplePredictor class generates random but realistic football scores:

• Common scores like 1:0, 2:1, 1:1, 3:1, etc.
• Based on the Python reference implementation patterns

Form Field Detection

• Automatically finds betting input fields ending with _heimTipp and _gastTipp
• Handles hidden form fields and submit buttons correctly
• Follows the same patterns as the Python kicktipp-cli

Safety Features

• Dry run mode: Shows what would be bet without actually submitting
• Existing bet detection: Skips matches where bets are already placed
• Override option: Can override existing bets if needed (implemented but not exposed in UI)
• Interactive confirmation: User must explicitly confirm bet placement

Dependencies

• .NET 9.0: Modern C# runtime
• AngleSharp: HTML parsing and DOM manipulation
• DotNetEnv: Environment variable loading from .env files

Project Structure

dev/Poc/
├── Program.cs              # Main application entry point
├── Services/
│   └── KicktippService.cs  # Core web automation logic
├── Models/
│   └── KicktippModels.cs   # Data models and predictor logic
└── .env.example            # Environment template (instructions only)

KicktippAi.Secrets/         # External secrets directory
└── dev/Poc/
    ├── .env.example        # Copy of environment template  
    └── .env                # Your actual credentials (gitignored)

Future Enhancements

Potential next steps:

• Multi-community support (beyond hardcoded "ehonda-test")
• More sophisticated prediction algorithms
• Command-line arguments for configuration
• Scheduling and automation features
• Better error handling and retry logic

Acknowledgments

This project is inspired by and follows the patterns established in schwalle/kicktipp-betbot, the original Python implementation for kicktipp.de automation. We specifically used the ehonda/kicktipp-cli fork (dev branch) as a reference for translating the Python implementation concepts to modern C#.