sunk-core-dump

Name: sunk-core-dump
Availability: InStock
Author: CoreWeave

Generate a comprehensive diagnostic dump of SUNK cluster resources and state.

7|Updated Jul 9, 2026

Search Tags:#troubleshooting #Kubernetes #Slurm #SUNK #diagnostic dump

Ready to run

npx skills add https://github.com/coreweave/sunk-anywhere --skill sunk-core-dump

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: sunk-core-dump
Source: https://github.com/coreweave/sunk-anywhere/tree/main/skills/universal/sunk-core-dump
Command: npx skills add https://github.com/coreweave/sunk-anywhere --skill sunk-core-dump

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires kubectl, helm, python3, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a complete diagnostic dump of a SUNK (Slurm on Kubernetes) cluster to assist with troubleshooting and debugging.

Core Features & Use Cases

Complete Cluster Dump: Collects Kubernetes resources, Slurm state, pod logs, GPU health, cgroup diagnostics, network/storage status, and SUNK CRDs.
Timestamped Output: Saves output to a timestamped directory for easy organization and access by support teams.
Use Case: When troubleshooting a SUNK cluster, use this Skill to gather a comprehensive set of diagnostic information to share with support.

Quick Start

Use the sunk-core-dump skill to generate a diagnostic dump of your SUNK cluster and save it to a local directory.

Frequently Asked Questions about sunk-core-dump

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

FAQPage Schema

How do I collect a comprehensive diagnostic dump for a SUNK cluster?▼

To collect a SUNK cluster diagnostic dump, you can generate a comprehensive report containing Kubernetes resources, Slurm state, pod logs, GPU health, cgroup diagnostics, network/storage status, and SUNK CRDs in a timestamped directory.

What information is included in a Slurm on Kubernetes cluster troubleshooting report?▼

A Slurm on Kubernetes troubleshooting report includes Kubernetes resources, Slurm state, pod logs, GPU health, cgroup diagnostics, network status, storage status, and SUNK CRDs to assist support teams.

Do I need kubectl and helm installed to capture SUNK cluster diagnostics?▼

Yes, generating a SUNK cluster diagnostic dump requires kubectl, helm, and python3 installed to successfully query Kubernetes resources, Slurm state, and execute diagnostic scripts.

How does generating a diagnostic dump help with Kubernetes troubleshooting?▼

Generating a diagnostic dump helps Kubernetes troubleshooting by consolidating pod logs, GPU health, cgroup diagnostics, and cluster state into a timestamped directory for support teams to analyze.

Can I capture Slurm state and GPU health simultaneously for cluster debugging?▼

Yes, you can capture Slurm state and GPU health simultaneously by running a comprehensive diagnostic dump that aggregates Kubernetes resources, cgroup diagnostics, and network/storage status together.

What is the best way to gather SUNK CRDs and pod logs for support analysis?▼

The best way to gather SUNK CRDs and pod logs is to generate a complete cluster diagnostic dump, which automatically saves all extracted state information and logs to a local timestamped directory for support.