take-me-cloud

take-me-cloud is a CLI for working with Lightning AI studios from your shell. It can list every studio you can access, sync Lightning SSH entries into ~/.ssh/config, create or replace studios from a local config file, and bootstrap a fresh studio for repository work.

Introduction

The CLI authenticates with Lightning AI through LIGHTNING_API_KEY and either LIGHTNING_USER_ID or LIGHTNING_USERNAME. It uses those credentials to discover your organizations, teamspaces, and studios.

Installation

pip install take_me_cloud

uv tool install take_me_cloud

For development:

pip install -e .

Configuration

Create ~/.config/take_me_cloud_config.yaml with a Lightning AI section:

Lightning AI:
  machine_default: T4
  cloud_provider: AWS
  teamspace:
    - name: rwth-gut/skillcomp
      owner_type: org
    - name: vaishnavahari/myml
      owner_type: user
      machine_default: CPU

teamspace entries are shown interactively when you create a studio. Each entry must use owner/teamspace naming.

Usage

List studios

take-me-cloud --list
take-me-cloud -ls

The output is grouped by teamspace and shows the studio name, owner, machine type, and current state.

Sync SSH config

take-me-cloud --lock-ssh

This preserves non-Lightning SSH hosts, removes stale Lightning entries, and writes the current accessible studios into ~/.ssh/config.

Create or replace a studio

take-me-cloud --create-replace my-studio

The CLI prompts for a teamspace from the config file, deletes any existing studio with the same name, starts the replacement with the configured machine type, and appends common setup commands to ~/.bash_history.

Go studio

take-me-cloud --go-studio my-repo

This creates or replaces a studio, clones svaichu/my-repo into the studio home directory, writes .vscode/settings.json, and installs the requested VS Code extensions on the remote environment.

Version

take-me-cloud --version
take-me-cloud -v

Development

Run the tests with:

uv run python -m unittest discover -s tests -v

Project Layout

• take_me_cloud/base.py: Lightning helpers and SSH/config logic
• take_me_cloud/cli.py: Argument parsing and command dispatch
• tests/: Unit tests for the core flows