⚡ MCP Energy Hub

Real-Time Energy Grid Intelligence for Carbon-Aware AI

Enterprise-grade MCP server providing real-time US power grid intelligence for carbon-aware AI compute scheduling

📖 Documentation • 🚀 Quick Start • 🤝 Contributing • 📜 License

🎯 The Problem

AI compute is exploding, but the grid isn't always green.

Data centers consume 1-2% of global electricity and growing rapidly
AI training runs can emit as much CO2 as 5 cars over their lifetime
Most AI workloads run without awareness of grid carbon intensity
Enterprises lack tools to schedule compute when renewables are high

💡 The Solution

MCP Energy Hub is an enterprise-grade MCP server that gives AI agents real-time visibility into the US power grid, enabling carbon-aware compute scheduling.

┌─────────────────────────────────────────────────────────────────┐
│                    🤖 AI Agent (Claude, etc.)                    │
│                              │                                   │
│                    ┌─────────▼─────────┐                        │
│                    │   MCP Protocol    │                        │
│                    └─────────┬─────────┘                        │
│                              │                                   │
│              ┌───────────────▼───────────────┐                  │
│              │      ⚡ MCP Energy Hub         │                  │
│              │  ┌─────────────────────────┐  │                  │
│              │  │ 8 MCP Tools for Energy  │  │                  │
│              │  │ Grid Intelligence       │  │                  │
│              │  └─────────────────────────┘  │                  │
│              └───────────────┬───────────────┘                  │
│                              │                                   │
│    ┌─────────────┬───────────┼───────────┬─────────────┐       │
│    ▼             ▼           ▼           ▼             ▼       │
│ ┌─────┐     ┌─────┐     ┌─────┐     ┌─────┐     ┌─────┐       │
│ │ERCOT│     │CAISO│     │ PJM │     │NYISO│     │MISO │       │
│ │Texas│     │Calif│     │ Mid │     │ NY  │     │Midwest      │
│ └─────┘     └─────┘     └─────┘     └─────┘     └─────┘       │
└─────────────────────────────────────────────────────────────────┘

✨ Key Features

Feature	Description
🌍 7 Grid Regions	ERCOT, CAISO, PJM, NYISO, MISO, SPP, ISONE
⚡ Real-Time Data	Live from EIA (US Energy Information Administration)
🌱 Carbon Intensity	kg CO2/MWh for each region, updated hourly
🔋 Generation Mix	Natural gas, coal, nuclear, wind, solar, hydro
🏢 Data Center Tracking	Energy estimates, PUE, AI workload impact
🎯 Smart Scheduling	Find the greenest region for your compute
📊 AI Impact KPIs	Track AI's share of grid load
🔌 MCP Native	Full Model Context Protocol support

🛠️ MCP Tools

8 Tools for Energy Intelligence

Tool	Description	Use Case
`get_grid_realtime`	Real-time grid metrics	Monitor current load & generation
`get_grid_carbon`	Carbon intensity + recommendation	Carbon-aware scheduling
`get_grid_forecast`	Load & carbon forecast	Plan future workloads
`list_grid_regions`	Available grid regions	Discover coverage
`get_data_centers`	Data center info	Track facilities
`get_data_center_energy`	Energy consumption estimates	Audit energy use
`get_ai_impact`	AI compute KPIs	Measure AI's grid footprint
`get_best_region_for_compute`	Find greenest region	Optimize for carbon/cost

Example: Carbon-Aware Scheduling

# AI Agent asks: "Where should I run this training job?"

result = mcp.call_tool("get_best_region_for_compute", {
    "optimize_for": "carbon"
})

# Response:
{
    "recommendation": "CAISO",
    "reason": "Lowest carbon intensity at 180 kg CO2/MWh",
    "rankings": [
        {"region": "CAISO", "carbon": 180, "renewable_pct": 45},
        {"region": "ERCOT", "carbon": 320, "renewable_pct": 28},
        {"region": "PJM", "carbon": 420, "renewable_pct": 12}
    ]
}

🚀 Quick Start

Prerequisites

Python 3.11+
EIA API Key (free)

Installation

# Clone the repository
git clone https://github.com/your-username/mcp-energy-hub.git
cd mcp-energy-hub

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Configure environment
cp .env.example .env
# Edit .env and add your EIA_API_KEY

Run the Server

# Start the FastAPI server
python -m uvicorn app.main:app --reload --port 8000

# Or run the standalone MCP server (for Claude Desktop)
python mcp_server.py

Try the API

# Get carbon intensity for Texas grid
curl -X POST http://localhost:8000/mcp/tools/call \
  -H "Content-Type: application/json" \
  -d '{"name": "get_grid_carbon", "arguments": {"region_id": "ERCOT"}}'

Example Response

{
  "success": true,
  "result": {
    "region_id": "ERCOT",
    "timestamp": "2024-11-28T22:00:00Z",
    "carbon_intensity_kg_per_mwh": 320.5,
    "renewable_fraction_pct": 28.3,
    "recommendation": "Good - Moderate carbon intensity"
  }
}

Connect to Claude Desktop

Add to your Claude Desktop MCP settings (claude_desktop_config.json):

{
  "mcpServers": {
    "energy-hub": {
      "command": "python",
      "args": ["/absolute/path/to/mcp-energy-hub/mcp_server.py"],
      "env": {
        "EIA_API_KEY": "your-api-key-here"
      }
    }
  }
}

📊 API Endpoints

Endpoint	Method	Description
`/docs`	GET	Interactive Swagger UI
`/mcp/info`	GET	MCP server information
`/mcp/tools`	GET	List all MCP tools
`/mcp/tools/call`	POST	Execute an MCP tool
`/grid/regions`	GET	List grid regions
`/grid/{region}/realtime`	GET	Real-time metrics
`/grid/{region}/carbon`	GET	Carbon intensity
`/health`	GET	Health check

🏗️ Architecture

┌────────────────────────────────────────────────────────────┐
│                     MCP Energy Hub                          │
├────────────────────────────────────────────────────────────┤
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────────┐ │
│  │   FastAPI    │  │  MCP Server  │  │  Data Ingestion  │ │
│  │   REST API   │  │  (8 Tools)   │  │  (EIA Collector) │ │
│  └──────┬───────┘  └──────┬───────┘  └────────┬─────────┘ │
│         │                 │                    │           │
│         └─────────────────┼────────────────────┘           │
│                           │                                │
│                    ┌──────▼──────┐                        │
│                    │  SQLite DB  │                        │
│                    │ Grid Metrics│                        │
│                    └─────────────┘                        │
├────────────────────────────────────────────────────────────┤
│                    External Data Sources                   │
│  ┌─────────┐  ┌─────────┐  ┌─────────┐  ┌─────────┐      │
│  │   EIA   │  │  ERCOT  │  │  CAISO  │  │   PJM   │      │
│  │   API   │  │   API   │  │   API   │  │   API   │      │
│  └─────────┘  └─────────┘  └─────────┘  └─────────┘      │
└────────────────────────────────────────────────────────────┘

🌍 Real-World Impact

For Enterprises

Reduce carbon footprint by scheduling AI workloads during high-renewable periods
Cost optimization by running compute when energy prices are low
ESG reporting with accurate AI energy consumption data

For AI Developers

Carbon-aware training - Train models when the grid is green
Transparent impact - Know your model's carbon footprint
Automated scheduling - Let AI agents make green decisions

Potential Impact

If 10% of AI workloads shifted to low-carbon periods: ~500,000 tons CO2/year saved
Real-time visibility enables 30-50% carbon reduction for flexible workloads

🔧 Tech Stack

Component	Technology
Backend	FastAPI, Python 3.11
Database	SQLite (HF) / PostgreSQL (Production)
MCP Protocol	Native implementation
Data Source	EIA Open Data API
Deployment	Docker, Hugging Face Spaces

📁 Project Structure

mcp-energy-hub/
├── app/
│   ├── main.py              # FastAPI application
│   ├── config.py            # Configuration
│   ├── api/routes/          # REST endpoints
│   ├── mcp/                  # MCP server implementation
│   │   ├── server.py        # MCP protocol handler
│   │   ├── tools.py         # Tool definitions
│   │   └── routes.py        # HTTP MCP endpoints
│   ├── ingestion/           # Data collectors
│   │   └── eia_collector.py # EIA API integration
│   └── models/              # Database models
├── mcp_server.py            # Standalone MCP server (stdio)
├── Dockerfile               # HuggingFace deployment
└── README.md                # This file

� Docker Deployment

# Build the Docker image
docker build -t mcp-energy-hub .

# Run the container
docker run -p 8000:8000 -e EIA_API_KEY=your-key mcp-energy-hub

🧪 Testing

# Run tests
pytest

# Run with coverage
pytest --cov=app --cov-report=html

🤝 Contributing

Contributions are welcome! Please see our Contributing Guidelines for details.

Fork the repository
Create a feature branch (git checkout -b feature/amazing-feature)
Commit your changes (git commit -m 'Add amazing feature')
Push to the branch (git push origin feature/amazing-feature)
Open a Pull Request

🙏 Acknowledgments

Anthropic - For creating the MCP protocol
EIA - For open energy data APIs
FastAPI - For the excellent web framework

📜 License

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

🔗 Links

� MCP Protocol: modelcontextprotocol.io
⚡ EIA Open Data: eia.gov/opendata
🐛 Report Issues: GitHub Issues

Made with ❤️ for sustainable AI

Helping AI compute become carbon-aware, one query at a time ⚡🌱

⭐ Star this repo if you find it useful!