# Kolada MCP Server [Health: Active]

**Category:** 🧮 Data Science Tools  
**Repository:** https://github.com/isakskogstad/kolada-mcp  
**GitHub Stars:** 12  
**npm Downloads (last month):** 252  
**Views:** 2  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/kolada-mcp-server

## Description
Swedish municipality statistics from Kolada API. 6000+ KPIs for all 290 municipalities.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "kolada-mcp-server": {
    "command": "npx",
    "args": ["-y","kolada-mcp-server"]
  }
}
```

## Documentation

## What Kolada MCP Server MCP server does

Kolada MCP Server MCP server exposes Swedish municipality and regional statistics through the Kolada API. The data covers more than 5,000 KPIs across 264 operating areas, including schools, healthcare, population, environment, economics, and democracy. Coverage includes all 290 Swedish municipalities and 21 regions.

The server also provides access to organizational units such as schools and elderly care facilities. It is developed independently from Kolada and RKA. Data use is subject to Kolada’s stated conditions, including source attribution when Kolada data is used in a service.

## How it works

The package runs as a local MCP process using stdio transport. An MCP client starts the process with `npx -y kolada-mcp-server`, after which the client can call the available tools. A global npm installation or a build from the repository can be used instead.

Kolada MCP Server MCP server separates catalog discovery from data retrieval. An agent can first search KPI names, groups, operating areas, municipalities, or organizational units, then use the returned identifiers to request metadata or values. Data tools support gender filtering with the codes T for total, M for men, and K for women.

Catalog responses use a 24-hour cache.

## Setup and configuration

The quickest installation is:

```bash
npx kolada-mcp-server
```

For a global installation, run `npm install -g kolada-mcp-server` and then start `kolada-mcp-server`. To build from source, clone the repository, install dependencies, run `npm run build`, and start `node dist/index.js`.

Claude Desktop, Claude Code, OpenAI Codex CLI, and Gemini CLI each have documented stdio configurations using the `npx` command. Other MCP clients that support stdio can use the same command and package arguments.

## Tools and capabilities

Kolada MCP Server MCP server includes tools for:

- Searching, inspecting, grouping, and filtering KPIs by operating area.
- Searching municipalities and regions, retrieving municipality details, and listing municipality groups.
- Finding organizational units, retrieving their details, and listing unit types.
- Fetching KPI data and all KPIs associated with a municipality.
- Comparing two to ten municipalities.
- Producing KPI time-series trends.
- Analyzing KPI values across municipalities with statistics and ranking.
- Filtering municipalities by KPI thresholds.
- Comparing KPIs for correlation.

The KPI retrieval tool supports up to 25 KPIs in one request. Listed organizational unit types include preschool, primary school, upper-secondary school, adapted school, adult education, elderly care, and leisure centers.

## Limitations and notes

The server depends on data available from the Kolada API. Kolada notes that data is published throughout the year according to the schedules of the underlying sources, and that values may be revised without separate notice.

Kolada data may be used commercially under the stated terms, but a service using unmodified Kolada data should identify Kolada as the source. The server must not be presented as an official partnership with RKA or Kolada. The repository describes the service as provided as-is and directs security reports to its security policy.

_Full upstream README: https://allmcps.com/mcp/kolada-mcp-server/readme_