Big Watermelon Daily Deals MCP Server

An MCP (Model Context Protocol) server that automatically fetches and analyzes daily produce deals from Big Watermelon, a fruit and vegetable wholesale store in Melbourne, Australia. The server uses Requesty.ai to extract deal information from images and makes it available to AI agents via the MCP protocol.

Features

• Automated Daily Fetching: Retrieves deals once per day after 7 AM Australia/Melbourne time
• AI-Powered Extraction: Uses Requesty.ai to analyze deal images and extract structured data
• Intelligent Caching: 24-hour cache to minimize API calls and improve performance
• Production-Ready: Includes rate limiting, graceful shutdown, comprehensive logging, and error handling
• Highly Configurable: All settings configurable via environment variables
• Performance Optimized: Concurrent processing with exponential backoff retry logic
• SSE Transport: Server-Sent Events for real-time MCP communication

Architecture

graph TB
    subgraph "MCP Clients"
        A[Cline/Roo Code]
        B[Claude Desktop]
        C[AnythingLLM]
    end
    
    subgraph "MCP Server"
        D[SSE Transport]
        E[Rate Limiter]
        F[MCP Handler]
    end
    
    subgraph "Core Logic"
        G[Deals Fetcher]
        H[Cache Manager]
        I[Image Processor]
    end
    
    subgraph "External Services"
        J[Big Watermelon Website]
        K[Requesty.ai API]
    end
    
    A --> D
    B --> D
    C --> D
    D --> E
    E --> F
    F --> G
    G --> H
    H --> I
    I --> J
    I --> K
    
    style D fill:#e1f5ff
    style G fill:#fff4e1
    style K fill:#ffe1e1

Requirements

• Go: 1.24.3 or later
• Requesty.ai API Key: Required for image analysis
• Optional: Docker for containerized deployment

Installation

Local Setup

1. Clone the repository:

   git clone https://github.com/hebra/ahemseepee/daily-deals-mcp-server.git
   cd daily-deals-mcp-server
   

2. Set up environment variables:

   export REQUESTY_API_KEY=your-requesty-api-key-here
   

3. Install dependencies:

   make deps
   

4. Build and run:

   make build
   make run
   

The server will start on http://localhost:8080 with the following endpoints:

• GET /sse - SSE connection endpoint for MCP clients
• POST /message - MCP message endpoint
• GET /health - Health check endpoint
• GET /ready - Readiness check endpoint

Docker Setup

# Build Docker image
make docker-build

# Run Docker container
make docker-run

Configuration

All configuration is done via environment variables with sensible defaults:

Required Configuration

Variable	Description	Default
REQUESTY_API_KEY	Requesty.ai API key for image analysis	(Required)

Requesty.ai Configuration

Variable	Description	Default
REQUESTY_BASE_URL	Requesty.ai API base URL	https://router.requesty.ai/v1
REQUESTY_MODEL	AI model to use	google/gemini-2.5-flash
REQUESTY_MAX_TOKENS	Maximum tokens in response	4096
REQUESTY_TEMPERATURE	Model temperature (0.0-1.0)	0.0

Server Configuration

Variable	Description	Default
PORT	HTTP server port	8080
RATE_LIMIT_REQUESTS	Maximum requests per window	100
RATE_LIMIT_WINDOW	Rate limit time window	1m

Fetching Configuration

Variable	Description	Default
FETCH_HOUR	Hour to fetch deals (0-23, Melbourne time)	7
CACHE_FILE	Path to cache file	bigwatermelon-dailydeals.cached.json
SPECIALS_URL	URL to scrape for deals	https://www.bigwatermelon.com.au/category/specials/
TIMEZONE	Timezone for scheduling	`Australia/Mel

... View full README on GitHub