cube-mcp-server

ankurCES/cube-mcp-server

3.1

If you are the rightful owner of cube-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Cube MCP Server is a sample repository designed to connect to Cube.dev pre-aggregations and make time series data accessible to large language models (LLMs).

Cube MCP Server

Sample Repo for Connecting to Cube.dev pre-aggregations and makes time series data available to the LLM.

Steps

Setup

Install uv

curl -LsSf https://astral.sh/uv/install.sh | sh

Verify Installation

uv --version

Clone the repo and activate the virtual environment

git clone https://github.com/ankurCES/cube-mcp-server
source .venv/bin/activate

Create the env file

cp sample.env.json env.json

Add the necessary configs and save.

Next run the below

uv add "mcp[cli]" pandas pyarrow matplotlib requests PyJWT

Add config for Claude Desktop

{
  "mcpServers": {
    "mix_server": {
      "command": "path/to/uv",
      "args": ["--directory", "path/to/cube_server", "run", "main.py"]
    }
  }
}

To Run

uv run main.py