Kaggle FAQs

Question 1

What is Kaggle MCP?

Accepted Answer

Kaggle MCP is a Model Context Protocol server that enables you to interact with the Kaggle API. It provides tools for searching and downloading datasets, and generates prompts for creating EDA notebooks.

Question 2

What can I do with Kaggle MCP?

Accepted Answer

You can use it to easily search Kaggle for datasets, download them to your local machine, and generate prompts for AI models to create Exploratory Data Analysis (EDA) notebooks, streamlining your data science workflow.

Question 3

How do I integrate Kaggle MCP with Claude?

Accepted Answer

Add the Kaggle MCP server configuration to your `claude_desktop_config.json` file in Claude's settings.  This tells Claude how to connect to the Kaggle MCP server.

Question 4

How do I install and run the Kaggle MCP server?

Accepted Answer

First, clone the repository, create a virtual environment, and install the dependencies using `pip install -r requirements.txt` or `uv sync`. Then, set up your Kaggle API credentials and run the server using `uv run kaggle-mcp`.

Question 5

What are the different ways to set up Kaggle API credentials?

Accepted Answer

You can use environment variables (recommended) by creating a `.env` file with your Kaggle username and API key. Alternatively, you can place your `kaggle.json` file in the expected location (usually `~/.kaggle/kaggle.json` on Linux/macOS).

Kaggle

概要

主な機能

ユースケース