Claude Code Skills for HPC

A collection of Claude Code skills for high-performance computing workflows, focusing on Slurm cluster job submission.

What are Claude Code Skills?

Skills are reusable prompt templates that extend Claude Code's capabilities. They live in .claude/skills/ and can be invoked with /skill-name.

Installation

1. Copy the skill directory to your project:

   cp -r submit-slurm-job /path/to/your/project/.claude/skills/

2. Configure paths in your project's .claude/CLAUDE.md:

   ## Slurm Configuration
   
   - PYTHON_PATH: `/path/to/miniconda3/envs/myenv/bin/python3`
   - PROJECT_DIR: `/home/your_user/private/homefile`   # scripts/logs must live here
   - PARTITION: `home`                                 # compute partition

3. Restart Claude Code or reload the project.

Available Skills

submit-slurm-job

Submit GPU compute jobs to Slurm clusters. Handles:

• Automatic sbatch script generation
• GPU type selection (V100, A100, A800, H200, etc.)
• Node scheduling
• Log file management
• Cluster rules: submit from ~/private/homefile and always specify GPU model in --gres

Usage:

/submit-slurm-job

Claude will ask you for job parameters (script path, GPU type, memory, etc.) and generate + submit the sbatch script.

Example:

User: /submit-slurm-job
Claude: What would you like to run?
User: Train my model with train.py --epochs 100
Claude: [generates and submits job]

Requirements

• Slurm cluster with GPU nodes
• Claude Code (formerly AWS Code)
• Python environment for your compute jobs

Contributing

Contributions welcome! Please ensure:

• Skills are well-documented
• Paths are parameterized (no hardcoded user paths)
• Examples are included

License

MIT