MCP ESIOS

A Python package that provides access to REE ESIOS API as MCP tools.

Installation

Clone the repository:

git clone https://github.com/yourusername/mcp-esios.git
cd mcp-esios

Install the package in development mode:

pip install -e .

Or using uv:

uv pip install -e .

Usage

Environment Variables

You must set the ESIOS_API_TOKEN environment variable with your ESIOS API token:

export ESIOS_API_TOKEN=your_api_token_here

Running with UV

# Run with UV package manager
uv run mcp-esios

# Increase verbosity
uv run mcp-esios -v

# Debug level verbosity
uv run mcp-esios -vv

Docker

Build the Docker image:

docker build -t mcp-esios .

Run the container:

# Run with environment variable for the API token
docker run -e ESIOS_API_TOKEN=your_api_token_here mcp-esios

Integrating with Claude Desktop

You can integrate this MCP server with Claude Desktop by adding a configuration to the Claude desktop config file:

Docker-based Integration

{
  "mcpServers": {
    "MCP_ESIOS": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "-e", "ESIOS_API_TOKEN", "mcp-esios"],
      "env": {
        "ESIOS_API_TOKEN": "your_api_token_here"
      }
    }
  }
}

UV-based Integration

{
    "mcpServers": {
        "MCP_ESIOS": {
            "command": "uv",
            "args": [
                "--directory",
                "/path/to/mcp-esios",
                "run",
                "mcp-esios"
            ],
            "env": {
                "ESIOS_API_TOKEN": "your_api_token_here"
            }
        }
    }
}

Add the configuration to Claude Desktop's config file:

~/Library/Application Support/Claude/claude_desktop_config.json

For more information about MCP integration, see the MCP Quickstart Guide.

Available Tools

This MCP server provides the following tools:

1. search_indicators - Search for indicators by name or description
2. get_indicator_data - Retrieve data for a specific indicator within a date range