What is k3d in the context of this skill?

k3d is a lightweight wrapper that runs k3s (a minimal Kubernetes distribution) inside Podman containers, allowing for fast and efficient local cluster management.

How does GPU support work?

The skill provides automated configuration for the NVIDIA device plugin, enabling Kubernetes pods to access host GPU resources for machine learning and compute tasks.

Can I run multiple Kubernetes clusters simultaneously?

Yes, the skill supports a multi-instance flag (--instance) that allows you to run and manage several independent clusters on different ports.

Can k3d pods interact with other local services?

Yes, clusters are automatically joined to the 'overthink' network, allowing pods to reach services like Ollama or Jupyter via standard DNS names.

Do I need to install kubectl separately?

While having kubectl is helpful, the skill provides a shell command that allows you to execute kubectl commands directly within the cluster context.

k3d Kubernetes Manager

Name: k3d Kubernetes Manager
Author: atrawog

byatrawog

0•

Deployment & DevOps

Orchestrates lightweight k3s Kubernetes clusters in Podman containers with native GPU support and service discovery.

The k3d skill enables Claude to manage lightweight k3s clusters running within Podman containers on the Overthink network. It simplifies the complex task of local Kubernetes orchestration by providing commands for cluster creation, multi-instance management, and lifecycle control. A standout feature is its automated GPU passthrough for NVIDIA workloads, making it ideal for AI/ML development. By integrating directly with the Overthink network, it allows seamless DNS-based communication between Kubernetes pods and other local services like Ollama and JupyterLab, providing a robust environment for cloud-native testing and deployment.

Key Features

010 GitHub stars

02Simplified kubectl execution via integrated shell commands

03Rapid k3s cluster creation and lifecycle management using Podman

04Automated NVIDIA GPU passthrough for hardware-accelerated workloads

05Native integration with the Overthink network for cross-service discovery

06Support for multi-instance clusters with isolated API ports

Use Cases

01Simulating multi-node cluster architectures for DevOps and CI/CD workflow testing

02Running GPU-accelerated ML models within a containerized Kubernetes environment

03Developing and testing Kubernetes-based applications locally with full API compatibility

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add atrawog/overthink-plugins k3d

For use in Claude.ai and ChatGPT