Skip to content
Get packs

Lambda Labs GPU Cloud

Guides an agent to launch, manage and connect to Lambda Labs GPU cloud instances for ML work.

Source Download skill
Repository
Orchestra-Research/AI-Research-SKILLs
Stars
14,195
Licence
MIT
Files
3 files

What it does

Teaches the agent how to use Lambda Labs, a paid cloud service that rents powerful graphics processors (GPUs) by the hour for training and running AI models. Once installed, the agent knows the available GPU types and prices, how to create an account and SSH key, how to launch, list and terminate instances through the Python client or curl commands, and how to use persistent storage that survives restarts. It does not itself rent or pay for anything; it only gives the agent the know-how to call Lambda's API with your credentials.

Great for

  • ML engineers who need dedicated GPUs with full SSH access for long training runs
  • Teams that want a pre-configured stack (PyTorch, CUDA, NCCL) and multi-node Slurm clusters
  • Developers who want an agent to script launching and shutting down GPU instances

Skip it if

  • Anyone who wants to avoid spending money: launching instances bills per minute, and a forgotten instance keeps charging, so the agent should be checked before it creates or terminates resources
  • Users wanting serverless or spot-priced GPUs, since the skill itself points to Modal, RunPod or Vast.ai for those cases
  • Handling sensitive data on shared infrastructure without reviewing Lambda's own security and data policies, which this file does not cover

Try asking

  • “Show me the available GPU types and their hourly prices on Lambda Labs”
  • “Launch a 1x H100 instance in us-west-1 with my SSH key and attach my training filesystem”
  • “List my running Lambda instances and tell me their IPs and status”
  • “Terminate the instance named training-job once my checkpoints are saved”
Step-by-step guide: how to use Lambda Labs GPU CloudInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    lambda-labs.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip lambda-labs.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip lambda-labs.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Set a spending check and make sure the agent confirms instance type, region and cost before launching, and confirms termination so no instance is left running
  • Attach a persistent filesystem at launch time, since filesystems cannot be added later and save checkpoints there rather than on local disk
  • Keep the API key in an environment variable rather than pasting it into chat so it is not stored in conversation history

What's inside

  • references/advanced-usage.md
  • references/troubleshooting.md
  • SKILL.md

Redistributed unmodified from Orchestra-Research/AI-Research-SKILLs under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.