ckan-mcp-server by mjanez - MCP Server

CKAN MCP Server

Smart MCP Agent for CKAN Open Data Portals

Semantic search, direct SQL queries, geospatial analysis, and advanced exploration of CKAN catalogs.

[!TIP] Recommended Test Portal: https://catalogo.datosabiertos.miteco.gob.es/catalogo

Key Features

Advanced semantic search with SOLR filters
Automatic detection of geospatial datasets (SHP, GeoJSON, WMS/WFS, etc.)
SQL queries on DataStore resources
Automatic inspection of portal schemas and capabilities
Data preview in Markdown table format
Smart caching and automatic retries for maximum stability
Full support for MCP resources (ckan://...)

Quick Installation

git clone https://github.com/mjanez/ckan-mcp-server.git
cd ckan-mcp-server
uv venv
source .venv/bin/activate    # Windows: .venv\Scripts\activate
uv pip install -e .

Optional Configuration (.env)

CKAN_URL=https://catalogo.datosabiertos.miteco.gob.es/catalogo
CKAN_API_KEY=                  # Only if you need authenticated actions
LOG_LEVEL=INFO

Usage

Development Mode (with Web Inspector)

uv run fastmcp dev src/server.py

The MCP Inspector will automatically open in your browser to test all tools.

Production Mode

uv run fastmcp run src/server.py

Claude Desktop Integration

Add one or more servers to the Claude Desktop configuration file:

{
  "mcpServers": {
    "ckan-miteco": {
      "command": "uv",
      "args": ["--directory", "/absolute/path/ckan-mcp-server", "run", "fastmcp", "run", "src/server.py"],
      "env": { "CKAN_URL": "https://catalogo.datosabiertos.miteco.gob.es/catalogo" }
    },
    "ckan-datosgob": {
      "command": "uv",
      "args": ["--directory", "/absolute/path/ckan-mcp-server", "run", "fastmcp", "run", "src/server.py"],
      "env": { "CKAN_URL": "https://datos.gob.es" }
    }
  }
}

Configuration paths:

macOS/Linux: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

Available Tools

Tool	Description
`search_datasets`	Advanced search with filters (formats, organization, geospatial only, etc.)
`get_datasets`	Complete dataset information
`query_datastore`	SQL queries on tabular resources
`list_organizations`	List of organizations with dataset counts
`get_capabilities`	Detects active extensions (DataStore, Spatial, etc.)

MCP Resources

ckan://dataset/{id} → Complete dataset JSON
ckan://dataset/{id}/schema → DataStore Schemas
ckan://info → General portal information

Usage Examples with Claude

What datasets are available regarding air quality?
Search only for geospatial cadastre datasets
From resource 12345-abcd, execute: SELECT * FROM table WHERE year > 2020 LIMIT 20
Which organizations publish the most datasets?

Project

Key Technologies:

FastMCP v2.13+
ckanapi v4.7+
Pydantic v2
gettext
uv as package manager

Internationalization

The project uses gettext:

# Configure language (default is Spanish)
export CKAN_MCP_LANGUAGE=en  # English
export CKAN_MCP_LANGUAGE=es  # Spanish

# Compile translations (only if you modify .po files)
python src/i18n/compile.py

See src/i18n/README.md for more details.

Roadmap

Geospatial search by bounding box
Static map previews
Full DCAT-AP support and SPARQL queries
Automatic statistics on DataStore resources
Official Docker image
Tests + CI/CD

License

Acknowledgments

FastMCP Team
CKAN project, datos.gob.es and the Open Data Spain community

Recommended CKAN Portals for Testing

<div align="center"> <strong>Find this useful?</strong> → <a href="https://github.com/mjanez/ckan-mcp-server">Leave a ⭐ on GitHub!</a> </div>