Ontario Data Catalogue MCP Server

A comprehensive Model Context Protocol (MCP) server for the Ontario Data Catalogue, built with FastMCP. This server provides access to the Ontario Data Catalogue through CKAN API endpoints, enabling AI assistants to access open government data.

📊 Features

This MCP server provides access to the Ontario Data Catalogue API including:

Dataset Search and Discovery

• Search datasets by keywords and filters
• Get detailed dataset information and metadata
• List datasets with customizable filtering options
• Search by specific tags or categories

Organization and Taxonomy

• List all organizations publishing datasets
• Browse available tags and categories
• Explore dataset groups and classifications

Resource Data Access

• Access resource data through the datastore API
• Search within resource data
• Apply filters to resource queries

🚀 Installation

1. Clone or download this repository
2. Install the required dependencies:

pip install -r requirements.txt

Dependencies:

• fastmcp>=2.8.0 - FastMCP framework for MCP server creation
• requests>=2.31.0 - HTTP library for API calls
• typing-extensions - Enhanced type hints support

📖 Usage

Running the Server

Start the MCP server:

python ont_data_mcp_server.py

The server will start and listen for MCP connections.

MCP Client Configuration

Add this server to your MCP client configuration (e.g., Claude Desktop):

{
  "mcpServers": {
    "ont-data": {
      "command": "python",
      "args": ["/path/to/ont_data_mcp_server.py"],
      "env": {
        "ONTARIO_DATA_API_KEY": ""
      }
    }
  }
}

If you're using Claude in the browser, you can add this configuration in the MCP settings panel.

For Windsurf, add this to your ~/.config/windsurf/mcp_config.json file.

API Reference

• Ontario Data Catalogue
• CKAN API Documentation