What It Does
Mcore Run On Slurm packages a focused devops & cloud workflow for an AI agent. How to launch distributed Megatron-LM training jobs on a SLURM cluster. Covers a minimal sbatch skeleton, environment-variable setup for torch.distributed.run, CUDA_DEVICE_MAX_CONNECTIONS rules across hardware and parallelism modes, container conventions, monitoring, and per-rank failure diagnosis. It is best suited to users who can review the resulting actions and provide only the accounts, files, or command access needed for the task. It is not a substitute for human approval on destructive, financial, security-sensitive, or public-facing actions.
Popular Use Cases
- Inspect deployment or infrastructure state
- Automate a repeatable operations task
- Troubleshoot configuration and runtime issues
- Apply documented production practices
Install this Agent Skill
Claude Code
npx skills add https://github.com/NVIDIA/skills --skill mcore-run-on-slurmCommands derived from the public GitHub SKILL.md record. Checked 2026-10-18. Review the source before running them.
Requirements
| Claude Code | Required / review |
| Codex | Required / review |
| Public SKILL.md source | Required / review |
| Review instructions and requested permissions before installation | Required / review |
| Paid service | Check source |
| Supported system | Check source |
Why It’s Popular
Mcore Run On Slurm is a verified Agent Skill from NVIDIA with a public SKILL.md, compatible with Claude Code, Codex.
