Kolada MCP

🚀 Kolada MCP Server

Kolada MCP Server enables seamless integration between large language models (LLMs) and the comprehensive Swedish municipal and regional statistical database Kolada, providing structured access to thousands of key performance indicators (KPIs) for in - depth data - driven analysis.

🚀 Quick Start

Kolada MCP Server contains pre - cached KPI metadata. Delete kpi_embeddings.npz to refresh.

✨ Features

• Semantic Search: Query KPIs using natural language.
• Category Filtering: Access KPIs by thematic areas.
• Municipal and Regional Data Retrieval: Obtain KPI data or historical time series.
• Multi - year Comparative Analysis: Evaluate changes in KPI performance across different municipalities.
• Cross - KPI Correlation Analysis: Analyze relationships between KPIs.

📦 Installation

Install Kolada MCP dependencies using uv:

uv sync

💻 Usage Examples

Basic Usage

To run the server locally for development:

uv run mcp dev server.py

Open http://localhost:5173 in MCP Inspector for testing and debugging.

Advanced Usage

Claude Desktop Integration

Edit your claude_desktop_config.json file to add Kolada MCP Server:

Docker Image (Locally Built)

"KoladaDocker": {
  "args": [
    "run",
    "-i",
    "--rm",
    "--name",
    "kolada-mcp-managed",
    "kolada-mcp:local"
  ],
  "command": "docker",
  "env": {}
}

Pre - built Container via PyPI

"KoladaPyPI": {
  "args": ["kolada-mcp"],
  "command": "/Users/hugi/.cargo/bin/uvx"
}

Local UV Execution (No Docker)

Replace [path to kolada-mcp] with the actual path:

"KoladaLocal": {
  "args": [
    "run",
    "-w",
    "[path to kolada-mcp]",
    "mcp"
  ],
  "command": "uv"
}

📚 Documentation

Overview

Kolada MCP Server acts as an intelligent middleware between large language model applications and the Kolada database, facilitating easy querying and analysis of municipal and regional data. With semantic search capabilities and powerful analytical tools, Kolada MCP significantly simplifies the navigation and interpretation of the vast number of KPIs in Kolada.

Example Queries

Pose complex questions requiring data analysis to the Kolada MCP Server:

• Where in Sweden can one find places suitable for family living, with good schools and quality healthcare?
• Explore the link between unemployment and mental illness in the Västernorrland region.
• Identify the municipalities with the greatest improvement in preschool education quality over the past five years.
• Create a dashboard to visualize the municipalities with the best and worst public transportation.

Available Tools

1. list_operating_areas: Retrieve available KPI categories.
2. get_kpis_by_operating_area: List KPIs by category.
3. search_kpis: Discover KPIs through semantic search.
4. get_kpi_metadata: Obtain detailed KPI metadata.
5. fetch_kolada_data: Retrieve KPI values.
6. analyze_kpi_across_municipalities: Conduct in - depth analysis of municipal KPIs.
7. compare_kpis: Evaluate KPI correlations.
8. list_municipalities: List municipality IDs and names.

📄 License

The project is released under LICENSE.