mcpField
R
runpod-compute

runpod-compute

github.com/companion-incCloud & DevOps10.0k102 installsUpdated today
Repository

About runpod-compute

A skill for using the `runpodctl` CLI to create and manage persistent GPU pods for a specific research run. It is intended for experiments, benchmarks, and dataset-heavy work requiring SSH access.

Key features

  1. 01Create and manage persistent GPU pods
  2. 02Inspect available GPU types and prices
  3. 03Transfer files to and from pods
  4. 04Connect to pods over SSH

Use cases

  1. 01Run long-running research experiments
  2. 02Use large datasets for replication or benchmarks
  3. 03Resume multi-step research work on a pod

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/companion-inc/feynman --skill runpod-compute

After installing, type/runpod-computein Claude Code to confirm it loaded.

Details

From the project README

RunPod Compute

Use runpodctl CLI for persistent GPU pods with SSH access during a specific research run. Do not use this skill for provider administration outside that run; tie every pod to a replication, benchmark, or dataset-heavy research objective.

Setup

brew install runpod/runpodctl/runpodctl   # macOS
runpodctl config --apiKey=YOUR_KEY

Commands

CommandDescription
runpodctl create pod --gpuType "NVIDIA A100 80GB PCIe" --imageName "runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04" --name experimentCreate a pod
runpodctl get podList all pods
runpodctl stop pod <id>Stop (preserves volume)
runpodctl start pod <id>Resume a stopped pod
runpodctl remove pod <id>Terminate and delete
runpodctl gpu listList available GPU types and prices
runpodctl send <file>Transfer files to/from pods
runpodctl receive <code>Receive transferred files

SSH access

ssh root@<IP> -p <PORT> -i ~/.ssh/id_ed25519

Get connection details from runpodctl get pod <id>. Pods must expose port 22/tcp.

GPU types

NVIDIA GeForce RTX 4090, NVIDIA RTX A6000, NVIDIA A40, NVIDIA A100 80GB PCIe, NVIDIA H100 80GB HBM3

When to use

  • Long-running research experiments needing persistent state
  • Large research datasets required by a replication or benchmark
  • Multi-step research work with SSH access between iterations
  • Always stop or remove pods after experiments
  • Check availability: command -v runpodctl

Comments

Markdown supported