k8s-launch-kit-pipeline

Runs the NVIDIA Kubernetes networking deployment pipeline from cluster discovery through manifest generation and deployment.

14|5|Updated Nov 4, 2025

Search Tags:#kubernetes #cicd #manifest generation #cluster discovery #nvidia networking #sriov #spectrum-x

Ready to run

npx skills add https://github.com/NVIDIA/k8s-launch-kit --skill k8s-launch-kit-pipeline

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: k8s-launch-kit-pipeline
Source: https://github.com/NVIDIA/k8s-launch-kit/tree/main/skills/k8s-launch-kit-pipeline
Command: npx skills add https://github.com/NVIDIA/k8s-launch-kit --skill k8s-launch-kit-pipeline

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill eliminates the complexity of coordinating cluster hardware discovery, deployment profile selection, manifest generation, and Kubernetes deployment across NVIDIA networking environments.

Core Features & Use Cases

End-to-End Pipeline: Run discovery, profile resolution, manifest generation, and deployment in the correct order with a single l8k invocation.
Flexible Deployment Workflows: Support Ethernet, InfiniBand, Spectrum-X, SR-IOV, multirail, heterogeneous node groups, and custom operator namespaces.
Automation and Validation: Integrate with CI/CD using JSON output and non-interactive approval, or validate the complete workflow with dry-run mode.
Use Case: Deploy NVIDIA networking components to a newly provisioned Kubernetes cluster after discovering its NIC hardware and selecting the appropriate SR-IOV or Spectrum-X profile.

Quick Start

Ask the k8s-launch-kit-pipeline skill to discover the cluster, generate the appropriate NVIDIA networking manifests, and deploy them using a dry run first.

Frequently Asked Questions about k8s-launch-kit-pipeline

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

FAQPage Schema

How do I automate Kubernetes networking deployments for NVIDIA Spectrum-X and SR-IOV?▼

Automating Kubernetes networking deployments for Spectrum-X and SR-IOV involves running an end-to-end pipeline that handles cluster hardware discovery, profile resolution, manifest generation, and deployment in the correct order with a single invocation.

What is the best way to generate NVIDIA networking manifests for a heterogeneous Kubernetes cluster?▼

Generating NVIDIA networking manifests for a heterogeneous cluster requires discovering available NIC hardware first, then selecting the appropriate deployment profile to resolve and output the correct configuration manifests for deployment.

Can I integrate NVIDIA Kubernetes networking deployment into a CI/CD pipeline?▼

You can integrate NVIDIA Kubernetes networking deployment into a CI/CD pipeline by utilizing JSON output for machine-readable results and enabling non-interactive approval to automate the deployment safety confirmation step.

Does the NVIDIA Kubernetes networking pipeline support InfiniBand and Ethernet deployments?▼

The NVIDIA Kubernetes networking pipeline supports InfiniBand, Ethernet, multirail, and heterogeneous node group deployments, allowing flexible workflow configurations alongside custom operator namespaces.

How do I validate a Kubernetes networking deployment without applying changes to the cluster?▼

You can validate a Kubernetes networking deployment without applying changes by executing the pipeline in dry-run mode, which simulates the complete workflow from discovery through manifest generation safely.

Do I need a kubeconfig to run the Kubernetes networking discovery phase?▼

Kubernetes cluster access and a valid kubeconfig input are required to run the discovery phase, discover NIC hardware, and execute the ordered phase deployment pipeline successfully.