How do I build a new container from a definition file?

You can use the 'ujust apptainer build' command with the --image flag pointing to your .def file and the --tag flag specifying your output .sif filename.

Where are my Apptainer images stored by default?

Images pulled or managed through the bazzite-ai-pod system are typically stored in ~/.local/share/apptainer/.

Does this skill support both NVIDIA and AMD GPUs?

Yes, the skill automatically detects the host GPU and applies the appropriate flags (--nv for NVIDIA or --rocm for AMD) during container execution.

What is the difference between Apptainer and Docker?

Apptainer is designed specifically for High-Performance Computing (HPC). Unlike Docker, it runs containers as the user (no root required), uses single-file SIF images, and is compatible with shared file systems common in clusters.

Apptainer HPC Container Manager

Name: Apptainer HPC Container Manager
Author: atrawog

byatrawog

デプロイメントとDevOps

Manages Apptainer containers for HPC workloads with automatic GPU passthrough and SIF image handling.

概要

The Apptainer skill empowers Claude to handle high-performance computing (HPC) container workflows using the Apptainer (formerly Singularity) standard. It simplifies the lifecycle of SIF images, allowing users to pull from registries like Docker Hub and NVIDIA NGC, build custom images from definition files, and run workloads with automatic NVIDIA/AMD GPU detection. This skill is ideal for researchers and developers who need rootless, portable container solutions for scientific computing, AI training, and cluster-based environments.

主な機能

Build SIF images from standard Apptainer definition files
Interactive shell and command execution within rootless containers
Pull images from Docker Hub, NVIDIA NGC, and Sylabs Cloud
0 GitHub stars
Automatic NVIDIA and AMD GPU detection and passthrough
Integrated cache management for optimized storage usage

ユースケース

Deploying reproducible AI training environments on HPC clusters
Converting Docker-based development workflows to rootless SIF images
Running GPU-accelerated scientific simulations without root access

概要

主な機能

Build SIF images from standard Apptainer definition files
Interactive shell and command execution within rootless containers
Pull images from Docker Hub, NVIDIA NGC, and Sylabs Cloud
0 GitHub stars
Automatic NVIDIA and AMD GPU detection and passthrough
Integrated cache management for optimized storage usage

ユースケース

Deploying reproducible AI training environments on HPC clusters
Converting Docker-based development workflows to rootless SIF images
Running GPU-accelerated scientific simulations without root access