# Podatki Si [Health: Active]

**Category:** 📊 Data Platforms  
**Repository:** https://github.com/pipeworx-io/mcp-podatki-si  
**GitHub Stars:** 0  
**npm Downloads (last month):** 428  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/podatki-si

## Description
Slovenia Open Data (podatki.gov.si) CKAN MCP.

## Tools
Capabilities this server exposes over MCP:

- **search_datasets** — Search Slovenia Open Data (Slovenia) for datasets by keyword. Returns each dataset's id/name, title, organization, and its resources (each with a resource_id for query_resource).
- **dataset** — Show full metadata for one Slovenia Open Data dataset by id or name (from search_datasets), including all resources and their resource_ids.
- **query_resource** — Pull rows from a tabular Slovenia Open Data resource (CKAN DataStore) by resource_id (from a dataset's resources). Optional full-text q, limit, offset. Note: only resources with the DataStore enabled are queryable.

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

```json
"mcpServers": {
  "podatki-si": {
    "command": "npx",
    "args": ["-y","@pipeworx/mcp-podatki-si"]
  }
}
```

## Documentation

## What Podatki Si MCP server does

Podatki Si MCP server exposes Slovenia’s public data catalog at podatki.gov.si through the Model Context Protocol. It is intended for agents that need to locate Slovenian datasets, inspect their available files or resources, and retrieve records from queryable tabular data.

The server works with CKAN dataset metadata and CKAN DataStore resources. Search results include identifiers, names, titles, organizations, and resource details. Those resource details include the identifiers needed for subsequent data queries.

## How it works

A client can connect to the hosted MCP endpoint at `https://gateway.pipeworx.io/podatki-si/mcp`. That endpoint is served through the Pipeworx gateway and lists the Podatki Si tools alongside shared Pipeworx meta-tools. The gateway can also route broader questions through `ask_pipeworx`.

For an isolated connection, Podatki Si MCP server can run locally over standard input and output with the `@pipeworx/mcp-podatki-si` npm package. The local process exposes only this pack’s tools and does not add the gateway’s shared meta-tools. The hosted and local options use the same source and tool set.

The normal workflow is to search for a topic, select a dataset, inspect its resources, and then pass a suitable `resource_id` to the row-query tool. A resource must have CKAN DataStore enabled before rows can be retrieved.

## Setup and configuration

Add the hosted endpoint to an MCP client that supports remote MCP connections:

```json
{
  "mcpServers": {
    "podatki-si": {
      "url": "https://gateway.pipeworx.io/podatki-si/mcp"
    }
  }
}
```

For local stdio operation, configure the npm package as the command:

```json
{
  "mcpServers": {
    "podatki-si": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-podatki-si"]
    }
  }
}
```

The package can also be started directly with `npx -y @pipeworx/mcp-podatki-si`. The provided setup does not require environment variables or a Pipeworx account for the standalone mode.

## Tools and capabilities

- `search_datasets` finds Slovenian datasets by keyword and returns dataset and resource identifiers.
- `dataset` returns complete metadata for a dataset selected by ID or name, including its resources.
- `query_resource` retrieves rows from a CKAN DataStore resource. It accepts an optional full-text query, limit, and offset.

Podatki Si MCP server is therefore suitable for dataset discovery and targeted tabular retrieval. It does not make every catalog resource queryable: only resources with DataStore support can be used with `query_resource`.

## Limitations and notes

The server covers the tools listed for the Slovenia Open Data pack. When using the gateway endpoint, `tools/list` also includes shared Pipeworx tools, so the visible tool count is larger than the pack-specific list. The local stdio package avoids those shared tools.

A dataset may contain resources that are downloadable or descriptive but lack an enabled DataStore. Such resources can still appear in metadata, but they cannot be queried with `query_resource`. The dataset or resource identifier should come from the search or metadata results before making a row query.

The project is distributed under the MIT license.

_Full upstream README: https://allmcps.com/mcp/podatki-si/readme_