Watsonx Visualization MCP Server

A comprehensive Model Context Protocol (MCP) server for automated data visualization and analysis, designed for seamless integration with IBM Watsonx Orchestrate.

🎯 Overview

This MCP server provides intelligent data visualization tools that automatically:

• Select the most appropriate chart type based on data structure
• Generate comprehensive data analysis and insights
• Create visualizations in multiple formats (HTML, PNG, Word)
• Support 40+ chart types including advanced visualizations
• Provide statistical analysis, trend detection, and recommendations

✨ Features

Intelligent Chart Selection

Automatically determines the best visualization type based on:

• Data structure and dimensions
• Temporal patterns
• Hierarchical relationships
• Statistical distributions
• Correlation patterns

Supported Chart Types

• Basic: bar, column, line, area, pie
• Stacked: stacked bar, stacked column
• Statistical: boxplot, scatter, bubble, heatmap
• Hierarchical: sunburst, tree map, hierarchy bubble, packed bubble
• Temporal: line, area, dual axes lines, dual axes column
• Specialized: waterfall, gantt chart, radar, wordcloud, network, tornado
• KPI: kpi, bullet
• Tabular: table, crosstab
• Geographic: map, legacy map
• Advanced: marimekko, radial, spiral, decision tree

Comprehensive Analysis

• Statistical measures (mean, median, std dev, quartiles, etc.)
• Trend detection and forecasting indicators
• Correlation analysis
• Outlier detection
• Pattern recognition
• Actionable recommendations

Multiple Output Formats

• HTML: Interactive, responsive visualizations
• PNG: High-quality static images
• Word: Professional documents with analysis

🚀 Quick Start

Prerequisites

• Node.js 18+
• npm or yarn
• IBM Watsonx Orchestrate account

Installation

Note: The package is not yet published on npm. Use local installation:

1. Clone the repository:

git clone https://github.com/tdognin/watsonx-visualization-mcp.git
cd watsonx-visualization-mcp

2. Install dependencies:

npm install

3. Run tests to verify installation:

npm test

4. Start the MCP server:

npm start

📖 Detailed installation guide: INSTALLATION_LOCALE.md

🔧 Configuration

MCP Server Configuration

The MCP server needs to be configured in your system configuration file (not in this project).

📖 For detailed step-by-step instructions, see GUIDE_CONFIGURATION_MCP.md

Quick configuration example for ~/.config/mcp/settings.json:

{
  "mcpServers": {
    "watsonx-visualization": {
      "command": "node",
      "args": ["/FULL/PATH/TO/watsonx-visualization-mcp/src/mcp-server/index.js"],
      "env": {}
    }
  }
}

⚠️ Important: Replace /FULL/PATH/TO/ with the actual path to your project directory.

Watsonx Orchestrate Integration

🚀 Quick Setup Guide: SETUP_WATSONX_ORCHESTRATE.md - Step-by-step guide to connect your local MCP server to Watsonx Orchestrate

📖 Complete Documentation: WATSONX_INTEGRATION.md - Detailed integration guide with advanced options

📖 Usage

Tool 1: generate_visualization

Generate a single visualization with automatic chart type selection and analysis.

Parameters:

• data (required): JSON data to visualize
• chartType (optional): Specific chart type (auto-selected if not provided)
• outputFormat (optional): 'html', 'png', or 'word' (default: 'html')
• includeAnalysis (optional): Include data analysis (default: true)
• title (optional): Chart title
• options (optional): Additional chart configuration

Example:

{
  "data": [
    {"month": "Jan", "sales": 1200, "profit": 300},
    {"month": "Feb", "sales": 1500, "profit": 450},
    {"month": "Mar", "sales": 1800, "profit": 600}
  ],
  "title": "Q1 Sales Performance",
  "outputFormat": "html",
  "includeAnalysis": true
}

Tool 2: analyze_data

Perform comprehensive data analysis without visualization.

Parameters:

• data (required): JSON data to analyze
• analysisType (optional): 'statistical', 'trend', 'correlation', or 'summary'

Example:

{
  "data": [
    {"product": "A", "sales": 1200, "cost": 800},
    {"product": "B", "sales": 1500, "cost": 900}
  ],
  "analysisType": "statistical"
}

Tool 3: create_dashboard

Create a comprehensive dashboard with multiple visualizations.

Parameters:

• datasets (required): Array of dataset configurations
• outputFormat (optional): 'html' or 'word' (default: 'html')
• dashboardTitle (optional): Dashboard title

Example:

{
  "datasets": [
    {
      "data": [{"category": "A", "value": 100}],
      "chartType": "pie",
      "title": "Distribution"
    },
    {
      "data": [{"month": "Jan", "sales": 1200}],
      "chartType": "line",
      "title": "Trend"
    }
  ],
  "dashboardTitle": "Sales Dashboard",
  "outputFormat": "html"
}

📊 Data Format Examples

Simple Object Format

{
  "Category A": 100,
  "Category B": 200,
  "Category C": 150
}

Array of Objects Format

[
  {"category": "A", "value": 100, "target": 120},
  {"category": "B", "value": 200, "target": 180},
  {"category": "C", "value": 150, "target": 160}
]

Time Series Format

[
  {"date": "2024-01-01", "sales": 1200, "expenses": 800},
  {"date": "2024-02-01", "sales": 1500, "expenses": 900},
  {"date": "2024-03-01", "sales": 1800, "expenses": 1000}
]

Network Format

[
  {"source": "A", "target": "B", "value": 10},
  {"source": "B", "target": "C", "value": 20},
  {"source": "A", "target": "C", "value": 15}
]

Gantt Chart Format

[
  {"task": "Planning", "start": "2024-01-01", "end": "2024-01-15"},
  {"task": "Development", "start": "2024-01-16", "end": "2024-02-28"},
  {"task": "Testing", "start": "2024-03-01", "end": "2024-03-15"}
]

🏗️ Architecture

watsonx-visualization-mcp/
├── src/
│   ├── mcp-server/          # MCP server implementation
│   ├── visualization-engine/ # Chart generation
│   ├── analysis-engine/      # Data analysis
│   ├── output-generators/    # Format converters
│   └── utils/               # Helper utilities
├── docs/                    # Documentation
├── examples/                # Usage examples
├── tests/                   # Test suites
└── config/                  # Configuration files

🎨 Styling

All visualizations follow IBM Carbon Design System guidelines:

• IBM Plex Sans font family
• Carbon color palette
• Accessible color contrasts
• Responsive layouts
• Professional styling

🧪 Testing

Run tests:

npm test

Run with coverage:

npm run test:coverage

📚 Documentation

• API Reference - Complete API documentation
• Watsonx Integration Guide - Setup and integration instructions
• Examples Guide - How to use the examples and MCP server
• Usage Examples - Practical usage scenarios
• Contributing Guide - How to contribute to the project
• Changelog - Version history and updates

🤝 Contributing

Contributions are welcome! Please read our contributing guidelines before submitting PRs.

📄 License

MIT License - see LICENSE file for details

🆘 Support

For issues and questions:

• GitHub Issues: [Create an issue]
• Documentation: docs/
• Examples: examples/

🔄 Version History

v1.0.0 (Current)

• Initial release
• 40+ chart types supported
• Intelligent chart selection
• Comprehensive analysis engine
• Multiple output formats
• Watsonx Orchestrate integration

🙏 Acknowledgments

• Built with Model Context Protocol
• Visualizations powered by Plotly.js
• Styling based on IBM Carbon Design System
• Document generation with docx

---

Made with ❤️ for IBM Watsonx Orchestrate