A comprehensive system for converting AI-generated recipes to Paprika format, with seamless Claude Desktop integration.

Transform your recipe workflow from this:

1. Ask Claude to create a recipe
2. Copy the recipe text
3. Manually format and paste into Paprika

To this:

1. Ask Claude Desktop: "Create a Thai curry recipe and convert it to Paprika format"
2. Done! Recipe is automatically saved as a .paprikarecipes file ready for import

• 🤖 Claude Desktop Integration - Convert recipes without leaving Claude
• 📱 Command-Line Tool - Standalone converter for JSON files
• 🔄 Smart Format Handling - Automatically fixes array/string mismatches
• 📁 Batch Processing - Convert multiple recipes at once
• 🛡️ Robust Validation - Handles missing fields and data type issues
• 📱 Cross-Platform - Works on macOS, Windows, and Linux

1. Install Python 3.10+ and dependencies:

   pip install "mcp>=1.2.0"

2. Run the setup script:

   python setup_mcp.py

3. Restart Claude Desktop and start converting recipes!

1. Convert a single recipe:

   python json_to_paprika.py path/to/recipe.json

2. Convert all recipes in a directory:

   python json_to_paprika.py recipes/

3. Interactive mode:

   python json_to_paprika.py

• recipe_mcp_server.py - MCP server for Claude Desktop integration
• json_to_paprika.py - Command-line converter
• setup_mcp.py - Automated setup for Claude Desktop

• README_MCP.md - Detailed MCP setup and troubleshooting
• Sample recipes - Example JSON and converted .paprikarecipe files

• test_mcp.py - MCP server functionality tests
• test_array_conversion.py - Format validation tests

You: "Create a simple pasta recipe with what I have: tomatoes, garlic, olive oil, and basil"

Claude: [Creates detailed recipe]

You: "Convert this to Paprika format"

Claude: ✅ Recipe successfully converted to Paprika format!
📁 Output file: Simple_Pasta_with_Tomatoes.paprikarecipes
📍 Location: /Users/you/Documents/PaprikaRecipes/Simple_Pasta_with_Tomatoes.paprikarecipes

# Single file conversion
$ python json_to_paprika.py "Thai Green Curry.json"
Processing: Thai Green Curry.json
✅ Final recipe bundle created: output/Thai_Green_Curry_with_Chicken.paprikarecipes

# Batch conversion
$ python json_to_paprika.py recipes/
Processing: recipes/
Found: recipes/curry.json
Found: recipes/pasta.json
Found: recipes/cookies.json
✅ Final recipe bundle created: output/20240125_1430.paprikarecipes

• Python 3.10+ (use pyenv for version management)
• MCP 1.2.0+ (for Claude Desktop integration)
• Claude Desktop (for MCP features)
• Paprika Recipe Manager (for importing converted recipes)

The system accepts JSON recipes in this format:

{
  "name": "Recipe Name",
  "description": "Brief description",
  "ingredients": ["1 cup flour", "2 eggs", "1 tsp salt"],
  "directions": ["Mix ingredients", "Bake at 350°F", "Serve hot"],
  "prep_time": "15 minutes",
  "cook_time": "30 minutes",
  "servings": "4",
  "categories": ["Dinner", "Quick"]
}

Note: The system automatically handles:

• Array-to-string conversion for ingredients/directions
• Missing fields (filled with sensible defaults)
• Mixed data types and formatting inconsistencies

from recipe_mcp_server import ensure_recipe_fields, gzip_files_inline, zip_files_inline

# Validate and convert recipe data
recipe = ensure_recipe_fields(your_recipe_dict)

# Create Paprika file
gzipped_files = gzip_files_inline("/path/to/temp/dir")
paprika_file = zip_files_inline(gzipped_files)

Files are automatically saved to ~/Documents/PaprikaRecipes/ for easy access. Single recipes use the recipe name as filename, while batch conversions use timestamps.

MCP server not loading in Claude Desktop:

• Check Python version (3.10+ required)
• Verify pip install "mcp>=1.2.0" succeeded
• Restart Claude Desktop after running setup

Permission errors:

• Ensure ~/Documents is writable
• Try running with different user permissions

Array format issues:

• The system automatically converts arrays to strings
• No manual formatting needed

For detailed troubleshooting, see README_MCP.md.

• README_MCP.md - Complete MCP server setup guide
• Sample files - Example recipes in sample_files/
• Code comments - Inline documentation in all Python files

This project demonstrates integrating AI recipe generation with recipe management apps. Feel free to:

• Report issues or suggest improvements
• Add support for other recipe management formats
• Extend the MCP server with additional tools
• Improve the command-line interface

This project is licensed under the MIT License - see the LICENSE file for details.

What this means:

• ✅ Use for any purpose (personal, commercial, etc.)
• ✅ Modify and distribute freely
• ✅ No warranty or liability on my part
• ⚠️ Must include the original license and copyright notice

Transform your AI-generated recipes into perfectly formatted Paprika imports with zero manual work. 🍳✨