Florence MCP Servers

A comprehensive MCP server collection for building a personal sous chef ambient agent. This extensible architecture currently implements recipe search and management as the foundational server, with additional specialized servers planned for meal planning, pantry inventory, grocery delivery, and restaurant ordering.

🏗️ Architecture Overview

Florence is designed as a multi-server MCP ecosystem where each server handles a specific domain of food/cooking functionality. The servers work together through shared data models and can be deployed independently or as a complete suite.

🚀 Quick Start with Docker

1. Clone and setup:

   git clone https://github.com/craine-io/florence-mcp-servers.git
   cd florence-mcp-servers
   make setup

2. Add your API key:

   # Edit .env file with your Spoonacular API key
   nano .env

3. Start the services:

   make dev

4. Test it works:

   make test

📦 Current Implementation

• Recipe API Server: Search and manage recipes via Spoonacular API ✅
• PostgreSQL Database: Ready for meal planning and inventory data
• Redis Cache: Ready for performance optimization
• Development Tools: pgAdmin, Redis Commander (optional)

🗺️ Roadmap: Planned MCP Servers

Phase 1 (✅ Current)

• Recipe API Server: Recipe search, details, and nutrition analysis

Phase 2 (🚧 Planned)

• Meal Planning Server: Weekly/monthly meal planning with dietary preferences
• Pantry Inventory Server: Track ingredients, expiration dates, and shopping needs
• Restaurant Delivery Server: Uber Eats, DoorDash integration for ordering

Phase 3 (📋 Planned)

• Grocery Delivery Server: Instacart, Amazon Fresh integration
• Nutrition Tracking Server: Calorie tracking, dietary goal monitoring
• Calendar Integration Server: Sync meal plans with Google/Outlook calendars

Phase 4 (💡 Future)

• Equipment Management Server: Track kitchen equipment and maintenance
• Weather Integration Server: Adjust meal suggestions based on weather
• Health Integration Server: Connect with fitness trackers and health apps

Each server is designed to work independently or as part of the complete Florence ecosystem.

🛠️ Available Commands

make help              # Show all available commands
make dev               # Start development environment
make prod              # Start production environment
make test              # Run all tests
make logs              # View logs from all services
make clean             # Clean up all Docker resources

🔧 Development

Multi-Server Architecture

The project is structured for easy addition of new MCP servers:

servers/
├── recipe_api_server/     # ✅ Implemented
├── meal_planning_server/  # 🚧 Planned next  
├── pantry_server/         # 📋 Planned
└── [future servers]/      # 💡 Extensible

Local Development

# Work on individual servers
cd servers/recipe_api_server
pip install -r requirements.txt
cp .env.example .env
# Edit .env with API keys
python src/main.py

# Work on shared components
cd shared/
# Edit types/, utils/, config/

Docker Development

make dev               # Start all current services
make shell             # Access recipe server container
make logs-recipe       # View recipe server logs only

# Future: When more servers are added
make logs-meals        # Meal planning server logs
make logs-pantry       # Pantry server logs