MCP Server Vegalite

🚀 Model Context Protocol (MCP) Server for Data Visualization

This is an implementation of a Model Context Protocol (MCP) server that provides an interface for large language models to perform data visualization using Vega - Lite syntax.

🚀 Quick Start

✨ Features

The server offers two core tools:

save_data

• Function: Save a data table containing aggregated results to the server for subsequent visualization.
• Input:
  • name (string): The name of the data table to be saved.
  • data (array): An array of objects representing the data table.
• Return: A success message.

visualize_data

• Function: Visualize a data table using Vega - Lite syntax.
• Input:
  • data_name (string): The name of the data table to be visualized.
  • vegalite_specification (string): A JSON string representing the Vega - Lite specification.
• Return: If --output_type is set to text, it returns a success message and includes the complete Vega - Lite specification and data in the artifact key. If --output_type is set to png, it returns a base64 - encoded PNG image wrapped in an MPC ImageContent container.

💻 Usage Examples

Basic Usage

# Add the server to your claude_desktop_config.json
{
  "mcpServers": {
    "datavis": {
        "command": "uv",
        "args": [
            "--directory",
            "/absolute/path/to/mcp-datavis-server",
            "run",
            "mcp_server_datavis",
            "--output_type",
            "png" # or "text"
        ]
    }
  }
}