@aerugo
Basic information
An MCP server for Kolada.
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"kolada-mcp": {
"command": "python",
"args": [
"-m",
"kolada_mcp"
]
}
}
}List all KPI categories with counts
Get KPIs within a specific category
Semantic search for KPIs using natural language
Get detailed metadata for a specific KPI
Fetch raw KPI data for municipalities
Comparative analysis with rankings
Compare two KPIs (difference or correlation)
List municipalities/regions
Filter by KPI threshold
A standalone MCP server that provides AI applications with access to Sweden's Kolada municipal statistics API. It enables natural language queries against thousands of Key Performance Indicators covering Swedish public sector data. Designed for developers and data analysts integrating Kolada data into AI-driven workflows.
Install via pip (pip install -e .) or Docker (docker-compose up -d kolada-mcp). Run in stdio mode with kolada-mcp or in HTTP mode by setting MCP_TRANSPORT=http and PORT=8001. Integrate with Claude Desktop by adding a JSON entry to its MCP server configuration.
It offers 9 tools: listing operating areas, getting KPIs by area, semantic KPI search, KPI metadata, raw data fetching, comparative analysis across municipalities, comparing two KPIs, listing municipalities, and filtering municipalities by KPI threshold.
Python 3.11 or newer. The server has no external dependencies beyond the Kolada API and uses a standalone semantic search model.
It can be run directly via pip (stdio mode) or as a Docker container using the provided `docker-compose.yml`. The transport mode is configurable between stdio and HTTP.
All data is fetched live from Sweden's public Kolada municipal statistics API. No local data stores are used.
Environment variables: `MCP_TRANSPORT` (stdio or http), `PORT` (HTTP server port, default 8001), and `LOG_LEVEL` (default INFO).
See and control a real Windows PC you own — from any MCP client, locally or remotely. Observe (UIA + screenshots), click/type/drag/scroll, launch apps, owner Live View. BYOH: your machine, your key.
Painless E2E Automation for Mobile and Web
YC (S26) | Open Computer History | Continuously record your company computer work, map your workflows, help you find work worth automating, and power your agents' context
🚀 The fast, Pythonic way to build MCP servers and clients.
Yuppsy MCP: build and publish real estate websites from your AI assistant. Yuppsy lets your AI assistant create live websites for real estate agents and individual property listings. You don't need a designer or any code
Comments