Marine Agent: SWFL Coastal Intelligence 🌊

Phase 1: Stable Prototype Complete

A high-performance, mobile-responsive coastal intelligence platform and AI agent for SWFL paddlers and beachgoers. Integrates real-time government sensors (NOAA, NWS, FWC) and scientific reports (Mote BCRS) for safety and activity planning.

Phase 1 Achievements [COMPLETE]

• Unified Backend: FastAPI + MCP server with a single source of truth (GLOBAL_DATA_STORE).
• Real-Time Sync: Map markers and dashboards are 100% synchronized for all SWFL beaches.
• Dynamic Flag System: Official Florida Beach Flag logic (Green, Yellow, Red, Purple) integrated into map and UI.
• Activity Status Bar: Instant Green/Yellow/Red status for Paddling, Swimming, and Beach.
• Predictive Timing: Textual forecast parsing to provide "Best Window" timing tips.
• Mobile-First UX: Glassy, high-contrast UI optimized for iPhone Safari.

Roadmap (Phase 2)

• Radar Overlay: Real-time weather radar integration for the Coast Map.
• PWA Support: "Add to Home Screen" capability with manifest and service worker.
• Cloud Deployment: Production build for Vercel/Render.
• Advanced AI Agent: Expanding MCP tools for deeper astronomical and marine analysis.

Tech Stack

• Backend: Python 3.12, FastMCP, FastAPI
• Frontend: React, Vite, Mapbox GL, Lucide Icons
• Data Sources: NOAA CO-OPS, NWS Weather.gov, FWC-FWRI ArcGIS, Mote Marine Lab.

Getting Started

Installation

1. Clone the repository: ```bash git clone https://github.com/yourusername/marine-agent.git cd marine-agent ```

2. Create and activate a virtual environment: ```bash python3 -m venv venv source venv/bin/activate ```

3. Install dependencies: ```bash pip install -r requirements.txt ```

Usage with Gemini CLI

Add the server to your Gemini CLI configuration: ```bash gemini mcp add marine-agent --command "python3" --path "$(pwd)/marine_server.py" ```

License

MIT