Colab

Provides MCP actions to link, control, and monitor Google Colab sessions via a bridge server.

typescript

0

GitHub Stars

typescript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "devadalat-colab-mcp": {
      "command": "bun",
      "args": [
        "https://github.com/DevAdalat/colab-mcp/releases/download/v1.0.0/colab-bridge.js"
      ],
      "env": {
        "YOUR_TOKEN": "<YOUR_TOKEN>"
      }
    }
  }
}

You can control Google Colab instances through a high-performance MCP server that exposes actions like linking to Colab, executing shell or Python code, and monitoring system resources. This server is designed to be used with MCP clients to manage Colab remotely via a lightweight bridge.

How to use

You will use an MCP client to connect to this Colab MCP Server and issue actions such as connecting to a running Colab instance, running shell commands, executing persistent Python code, and viewing system information like CPU, GPU, and RAM usage. Start by launching the server with the provided runtime ( Bun ) option, then configure your MCP client to point to this server. Once connected, choose the available actions from your client to interact with Colab as if you were operating it locally.

How to install

Prerequisites: you need the Bun runtime installed on your machine.

Quick Start with no install required: run the server directly from its release URL using Bun.

Official Quick Start command you can copy and run in your shell:

bun https://github.com/DevAdalat/colab-mcp/releases/download/v1.0.0/colab-bridge.js

Additional setup for MCP clients

For a stable setup, you can download the bridge script locally and configure your MCP client to start it with Bun.

Download the bridge script (local path is referenced in the example):

curl -LO https://github.com/DevAdalat/colab-mcp/releases/download/v1.0.0/colab-bridge.js

Manual installation

If you prefer to run it locally, perform these steps in order.

Install dependencies ( Bun is required):

bun install

Run the server using the source file you have available (index.ts is the entry point in this setup, per the standard configuration):

bun run index.ts

Notes on requirements and environment

Ensure Bun is installed and that you have a running Google Colab instance with the bridge script active to accept MCP commands.

Available tools

connect_to_colab

Link your MCP client to a running Colab instance to establish a control session.

run_colab_shell

Execute shell commands on the linked Colab environment to perform system tasks or administrative operations.

run_colab_python

Run persistent Python code within the Colab session to perform data processing or experiments.

get_colab_system_info

Retrieve CPU, GPU, and RAM information to monitor resource usage and performance.