Keboola Platform Integration MCP Server

Data Science ToolsPython

Connect AI agents to Keboola data platform capabilities

Available Tools

storage_list_buckets

List all buckets in the Keboola project

storage_list_tables

List all tables in a bucket or the entire project

storage_table_detail

Get detailed information about a specific table

storage_table_preview

Preview data from a table with optional filtering and limiting

storage_update_table_description

Update the description of a table

storage_update_bucket_description

Update the description of a bucket

component_list

List all components (extractors, writers, etc.) in the project

component_detail

Get detailed information about a specific component

component_configuration_list

List all configurations for a specific component

component_configuration_detail

Get detailed information about a specific component configuration

job_run

Run a job for a specific component configuration

job_status

Check the status of a specific job

sql_query

Execute a SQL query in the Keboola workspace

sql_create_transformation

Create a SQL transformation using natural language

metadata_search

Search for metadata in the project using natural language

metadata_read

Read metadata for a specific object

metadata_update

Update metadata for a specific object

Keboola MCP Server provides a seamless bridge between your Keboola data platform projects and modern AI tools. It transforms Keboola's powerful features—including storage access, SQL transformations, and job triggers—into callable tools that can be used by AI assistants like Claude, Cursor, CrewAI, LangChain, and Amazon Q. This integration enables AI agents to directly query tables, manage data descriptions, create and run SQL transformations, and interact with project metadata using natural language. By eliminating the need for custom glue code, Keboola MCP Server delivers the right data to your AI agents exactly when and where they need it.

Introduction

Keboola MCP Server enables AI assistants to interact directly with your Keboola data platform. This integration allows AI tools to access data, run transformations, execute SQL queries, and trigger jobs without requiring custom integration code.

Prerequisites

Before setting up the Keboola MCP Server, ensure you have:

Python 3.10 or higher installed
Access to a Keboola project with admin rights
Your preferred MCP client (Claude, Cursor, etc.)
The uv package installer tool

Installing UV

The MCP client uses uv to automatically download and run the Keboola MCP Server.

macOS/Linux:

# Install using Homebrew
brew install uv

Windows:

# Using the installer script
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

# Or using pip
pip install uv

# Or using winget
winget install --id=astral-sh.uv -e

Required Credentials

You'll need three key pieces of information to set up the MCP server:

1. KBC_STORAGE_TOKEN

This authentication token grants access to your Keboola project. To create or manage tokens:

Log in to your Keboola project
Navigate to Users & Settings > API Tokens
Create a new token with appropriate permissions
- For full access, use the master token
- For limited access, create a custom storage token

2. KBC_WORKSPACE_SCHEMA

This identifies your workspace in Keboola and is required for SQL queries:

Go to your Keboola project
Navigate to SQL Workspace
Create a new workspace or use an existing one
Enable "Grant read-only access to all Project data" when creating the workspace
Note the workspace schema name

3. Keboola Region

Your Keboola API URL depends on your deployment region:

| Region | API URL | | --- | --- | | AWS North America | https://connection.keboola.com | | AWS Europe | https://connection.eu-central-1.keboola.com | | Google Cloud EU | https://connection.europe-west3.gcp.keboola.com | | Google Cloud US | https://connection.us-east4.gcp.keboola.com | | Azure EU | https://connection.north-europe.azure.keboola.com |

BigQuery-Specific Setup

If your Keboola project uses BigQuery backend, you'll need additional configuration:

Go to your Keboola BigQuery workspace and display its credentials (click Connect button)
Download the credentials file (JSON) to your local disk
Set the full path of the downloaded JSON credentials file to GOOGLE_APPLICATION_CREDENTIALS environment variable

Configuration

Claude Desktop Configuration

Go to Claude (top left corner) → Settings → Developer → Edit Config
Add the configuration from the installation section
Restart Claude desktop for changes to take effect

Config file locations:

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

Cursor Configuration

Go to Settings → MCP
Click "+ Add new global MCP Server"
Add the configuration from the installation section

Usage

Once configured, you can interact with your Keboola project through natural language in your AI assistant. For example:

"Show me the tables in my Keboola project"
"Run a SQL query to find the top 10 customers by revenue"
"Create a transformation to clean the customer data"
"Show me the status of my last job run"

The MCP server handles the translation between your requests and the Keboola API, providing a seamless experience for data exploration and manipulation.