How do I handle environment secrets?

You can register secrets in the RunPod web console and reference them in your runpod.toml file using the {{ RUNPOD_SECRET_XXX }} placeholder syntax.

Is there a way to test the configuration before launching?

Yes, both GPU and CPU creation scripts support a --dry-run flag that displays the command or request body without actually creating a billable instance.

Can I launch CPU-only instances with this skill?

Yes, the skill includes a dedicated script that uses the RunPod REST API to provision CPU pods, which is useful when GPU acceleration is not required.

What prerequisites are needed to use the RunPod skill?

You must have the runpodctl CLI installed and configured (~/.runpod/config.toml) and a runpod.toml file in your current working directory.

Does this skill support managing existing pods?

No, this skill is specifically designed for the creation and initial setup of pods. Tasks like stopping, removing, or listing existing instances must be handled through other tools or the RunPod console.

RunPod Pod Creator

Name: RunPod Pod Creator
Author: pokutuna

bypokutuna

•

클라우드 인프라

Automates the provisioning and configuration of RunPod GPU and CPU instances via local configuration files.

This skill streamlines the deployment of high-performance compute resources by allowing Claude to launch RunPod instances directly from your workspace. It handles both GPU-accelerated nodes for intensive AI/ML tasks and cost-effective CPU-only pods, leveraging a standardized runpod.toml configuration. By integrating the RunPod CLI and REST API, the skill manages the complexity of instance creation, SSH connectivity, and environment initialization, enabling developers to spin up on-demand infrastructure without leaving their coding environment.

주요 기능

01Automated creation of GPU and CPU pod instances from local config

02Template generation for runpod.toml and initialization scripts

031 GitHub stars

04Dry-run mode to validate configurations before incurring costs

05Support for custom GPU hardware overrides (e.g., RTX 5090)

06Instant SSH connection setup post-provisioning

사용 사례

01Provisioning lightweight CPU pods for background tasks or microservices

02Spinning up high-end GPU instances for machine learning model training

03Automating cloud environment setup for remote development workflows

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pokutuna/claude-plugins create-pod

For use in Claude.ai and ChatGPT

Download Skill