NVIDIANVIDIAOfficialยท4 Agent Skills Included

nvcf

Deploy and manage GPU cloud functions on Kubernetes

Installs, operates, and tears down self-hosted NVIDIA Cloud Functions control-plane and compute-plane stacks on Kubernetes. Covers cluster registration, Helmfile deployments, function and task lifecycle, and prerequisite setup like KAI Scheduler and SMB CSI. Removes guesswork from multi-cluster GPU deployments with step-by-step commands, safety checks, and troubleshooting references.
npx skills add NVIDIA/nvcf --all -g -y
Available:

Tells the agent how the monorepo is organized, how to route requests to the right subtree or skill, and what safety rules to follow before running destructive cluster or CI operations.

All Skills in This Repository (4)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install nvcf?โ–ผ

Run `npx skills add NVIDIA/nvcf --all -g -y` in your terminal to install all skills in this suite globally.

How to deploy GPU functions on my own Kubernetes cluster?โ–ผ

The nvcf-self-managed-cli and nvcf-self-managed-installation skills walk the agent through control-plane install, cluster registration, and function deploy and invoke using nvcf-cli and Helmfile.

What prerequisites does a self-hosted NVCF cluster need?โ–ผ

The prerequisite skill installs nvcf-cli, KAI Scheduler, and the SMB CSI driver with pinned, NVCF-validated Helm versions before the compute plane is installed.

Does nvcf work with Claude Code and Cursor?โ–ผ

Yes. All skills follow the universal SKILL.md standard and run in Claude Code, Cursor, Codex, and other compatible agents.

Can I safely tear down an NVCF deployment?โ–ผ

Yes. The skills enforce plan-only dry runs first and guide the agent through ordered compute-plane then control-plane teardown without touching unrelated cluster resources.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’