setup-cluster

Name: setup-cluster
Availability: InStock
Author: wangdepin

Configure SLURM, RunPod, or local GPU clusters in .raca/clusters.yaml.

Updated May 12, 2026

Search Tags:#authentication #configuration #runpod #slurm #cluster setup #gpu access #hpc validation

Ready to run

npx skills add https://github.com/wangdepin/raca-codex --skill setup-cluster-wangdepin

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: setup-cluster
Source: https://github.com/wangdepin/raca-codex/tree/main/.claude/skills/setup-cluster
Command: npx skills add https://github.com/wangdepin/raca-codex --skill setup-cluster-wangdepin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the problem of setting up and validating the connection between your compute environment and RACA so you can run experiments remotely without guesswork.

Core Features & Use Cases

Guided cluster onboarding: Walks you through setting up SLURM, RunPod, or local GPU backends with the required connection details.
Writes correct cluster configuration: Creates and updates .raca/clusters.yaml with the fields RACA needs for auth and job submission.
Verifies working connectivity and access: Checks SSH/RACA auth and, for SLURM, discovers partitions and validates actual GPU access using sbatch --test-only.
Use Case: When you want to run an experiment on torch at your university HPC, this Skill helps you configure the cluster, confirm you can submit GPU jobs, and set a safe default partition so your first run succeeds.

Quick Start

Use the setup-cluster skill to configure your SLURM HPC by telling it the cluster type and your connection details when prompted.

Frequently Asked Questions about setup-cluster

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I set up GPU access on a SLURM HPC cluster for remote experiment execution?▼

To set up SLURM HPC GPU access, you configure authenticated connectivity by providing connection details, which validates SSH access and checks partition GRES and account eligibility using sbatch --test-only.

Can I connect a RunPod cloud GPU backend to run remote experiments?▼

Yes, you can connect RunPod cloud GPUs by providing the required connection details during onboarding, which writes validated configuration to .raca/clusters.yaml and verifies operational access for remote execution.

What is the best way to validate cluster configuration before submitting GPU jobs?▼

The best way to validate cluster configuration is running operational checks that verify SSH and RACA auth, discover SLURM partitions, and confirm actual GPU access using sbatch --test-only to ensure your first run succeeds.

Does cluster setup support adding local GPU hosts alongside existing SLURM HPC backends?▼

Yes, cluster setup supports adding local GPU hosts alongside SLURM HPC and RunPod backends during initial onboarding and subsequent additions, writing validated configuration to .raca/clusters.yaml for each backend.

Why does my SLURM partition validation fail when configuring HPC cluster access?▼

SLURM partition validation fails when sbatch --test-only detects insufficient GRES or account eligibility, meaning your cluster configuration lacks the proper partition access required to submit GPU jobs.

Do I need to manually edit .raca/clusters.yaml to configure cluster authentication?▼

No, you do not need to manually edit .raca/clusters.yaml, because guided cluster onboarding automatically writes and updates the required connection fields for authentication and job submission.