Automation & Ops ✓ NVIDIA · Official kubernetesnemo-rlnrl-k8srayclusterrayjobtraining-monitoringhydra-overrides

NeMo-RL Kubernetes Launcher

Launch, monitor, iterate on, and debug NeMo-RL training runs on Kubernetes.

FollowSkills review · FSRS-2.0
Not recommended
54/ 100 5-point scale 2.7 / 5
Trust16 / 25 · 3.2/5

The skill explicitly requires checking shared-cluster state first and confirming before deleting shared infrastructure; secrets are referenced through secretKeyRef rather than embedded. Deductions apply because Bash, Write, and Edit are high-impact capabilities, the workflow includes cluster deletion, RayJob deletion, and job stopping, while confirmation semantics, authorization boundaries, rollback, and full data-flow disclosure remain incomplete. External wandb exposure is only indirectly addressed.

Reliability9 / 20 · 2.3/5

The document systematically covers commands, lifecycle modes, troubleshooting, and completion checks, with several diagnostic paths for failures. Static review did not execute the CLI, and the skill provides neither a committed key-path test suite nor a clear error-output contract. The benchmark is only a limited evaluation summary, so the score stays below the static ceiling of 10 and is reduced for unverified reproducibility and thin abnormal-input feedback.

Adaptability10 / 15 · 3.3/5

Audience, Kubernetes/NeMo-RL scenarios, trigger phrases, and ephemeral versus long-lived boundaries are fairly clear, including monitoring, debugging, and iteration workflows. Deductions apply because it depends on a specific NVIDIA/NeMo-RL repository layout, the nrl-k8s CLI, a Kubernetes cluster, and profile-specific infrastructure; non-fit and compatibility boundaries are limited, and Chinese-language support is not addressed. Core functionality is not entirely dependent on overseas services, although wandb and other external services may affect parts of the workflow.

Convention9 / 15 · 3.0/5

The SKILL.md has clear progressive sections, tables, command examples, gotchas, and a completion checklist; the skill card supplies Apache 2.0 licensing, owner, version 1.5.4, references, and evaluation information. Deductions apply because BENCHMARK explicitly reports missing metadata.author, metadata.tags, Instructions, and Examples. Maintenance ownership and update paths are only indirectly indicated by repository synchronization and publication context, and some prerequisites remain implicit.

Effectiveness6 / 15 · 2.0/5

The skill provides directly actionable workflows for launching, monitoring, retrieving logs, stopping, and tearing down runs. Positive evaluation tasks cover mode selection, driver-log retrieval, and prerequisite checks, indicating useful core coverage. Static review cannot verify command compatibility with the current CLI, cluster configuration, or APIs; users must still supply recipes, infra files, namespaces, permissions, and a working environment, so the score remains below the static ceiling of 7.

Verifiability4 / 10 · 2.0/5

The files provide concrete commands, paths, configuration fields, evaluation tasks, and a benchmark summary, creating some auditability. The skill directory lacks a committed test suite, CI evidence covering its key paths, or independent corroboration; benchmark execution details and the evidence chain are limited, so static review cannot verify the principal operational claims.

Evidence confidence:Low Reviewed Jul 20, 2026 Reviewed revision 55f18499943e
Before you use it
  • Before execution, confirm the target namespace, recipe/infra pairing, PVCs, Secrets, ServiceAccount, permissions, and shared-cluster ownership; cluster down, --replace, and RayJob deletion can affect others and have no automatic rollback.
  • Do not place wandb, HF, or image credentials in YAML or command-line arguments; verify that logs, project names, run names, and data paths do not expose sensitive information, and check external-service reachability.
  • The skill depends on a specific repository layout, nrl-k8s version, Kubernetes/KubeRay, scheduler, DRA, and hardware profile; validate commands against the target environment's CLI help, neighboring infra files, and authorization policy.
  • The BENCHMARK results are static publication material and do not replace independent validation on the target cluster; Chinese support, compatibility, and maintenance ownership are not sufficiently specified.
See the full review method →

What does this skill do, and when should you use it?

This skill is an operational playbook for running NeMo-RL recipes on a Kubernetes cluster through the nrl-k8s CLI. It covers ephemeral RayJob runs and reusable long-lived RayCluster workflows, with commands for validation, status inspection, log retrieval, stopping jobs, and cluster lifecycle management. It also explains Hydra overrides, shared-filesystem behavior, scheduling concerns, and troubleshooting for hung or failed training. It fits engineering teams that already have Kubernetes, KubeRay, Ray, and the surrounding NeMo-RL infrastructure available.

It instructs the agent to inspect the recipe, infrastructure files, git state, and Kubernetes prerequisites before acting. It runs nrl-k8s check, run, status, cluster, job, logs, and dev commands to render or apply RayJob, RayCluster, and Deployment resources, select upload, image, or lustre code sources, and monitor training. It uses kubectl, kubectl port-forward, curl, and Ray Dashboard APIs to inspect jobs, retrieve driver logs, stop runs, and verify RayJob teardown.

  1. A NeMo-RL engineer needs to submit a one-shot Kubernetes training run whose cluster should be removed after completion.
  2. A training developer needs a long-lived RayCluster for repeated recipe and parameter iterations.
  3. A platform operator needs per-role cluster, pod, daemon, and Ray Job status or terminal driver logs.
  4. A researcher has a stalled or failed run and needs to investigate GPU sizing, scheduling, DRA, shared storage, and logs.
  5. A team needs to test experiment changes with Hydra overrides without editing the shared recipe on the pod filesystem.

What are this skill's strengths and limitations?

Pros
  • Covers validation, submission, monitoring, stopping, debugging, and teardown workflows.
  • Clearly distinguishes ephemeral RayJob execution from long-lived RayCluster iteration.
  • Includes concrete guidance for GPU scheduling, DRA, shared filesystems, OmegaConf interpolation, and log retrieval.
  • Can be installed as an individual skill through the documented skills CLI flow.
Limitations
  • Requires an existing nrl-k8s CLI, kubectl, Kubernetes cluster, KubeRay, and Ray environment.
  • Operations affect shared clusters and potentially expensive GPU resources; an incorrect teardown or recreation can affect other users.
  • Specific hardware profiles are not fully defined in the skill; new profiles require reviewing neighboring infrastructure files.
  • It is a playbook rather than a standalone training executor, and the source does not provide skill-specific test results or a platform coverage matrix.

How do you install this skill?

Install the individual skill with the skills CLI command documented in the repository README: npx skills add nvidia/skills --skill launch-nemo-rl --yes. The README does not document installation of the nrl-k8s CLI or a Kubernetes cluster; those prerequisites must already be available.

How do you use this skill?

After installation, give the agent a concrete request such as “submit this NeMo-RL training job on Kubernetes” or “why is this NeMo-RL run stuck; inspect its logs.” Supply a recipe and matching infrastructure file, for example: nrl-k8s run <recipe> --infra <infra> --rayjob --dry-run. For a long-lived development loop, use nrl-k8s run <recipe> --infra <infra> --run-id <id>. Before applying infrastructure, verify the namespace PVC, Secrets, and service account; confirm before deleting shared resources.

FAQ

Does this skill create the underlying Kubernetes cluster?
It manages RayCluster and RayJob application, reuse, recreation, and teardown, but the source does not describe provisioning the underlying Kubernetes cluster. A target cluster must already exist.
When should I use ephemeral mode versus long-lived mode?
Use the default ephemeral mode for a one-shot run. Use --raycluster when the cluster should remain available for development and repeated iterations.
What should I inspect when a run is stuck?
Start with nrl-k8s status, Kubernetes RayJob/RayCluster and Pod state, then inspect GPU sizing, node selection, KAI scheduling, DRA claims, PVCs, Secrets, service accounts, and driver logs.
Does the skill have a stated cost?
The source gives no skill price. It does warn that actions on shared clusters and GPU infrastructure can have significant operational cost.

More skills from this repository

All from NVIDIA/skills

Dev & Engineering ✓ NVIDIA · Official

TAO Platform Execution SDK

Submit, monitor, and manage NVIDIA TAO GPU training jobs across supported platforms.

Dev & Engineering ✓ NVIDIA · Official

NeMo-RL Auto Research

Turns NeMo-RL research goals into reproducible, iterative experiments.

Writing & Content ✓ NVIDIA · Official

NeMo-RL Documentation Conventions

Keeps NeMo-RL documentation, indexes, and docstrings consistent.

Automation & Ops ✓ NVIDIA · Official

TAO Jobs on Kubernetes

Submit and monitor NVIDIA TAO GPU training jobs through Kubernetes.

Automation & Ops ✓ NVIDIA · Official

NeMo-RL Brev Runbook

Helps NeMo-RL agents manage storage, caches, logs, and authentication safely on Brev instances.

Automation & Ops ✓ NVIDIA · Official

DOCA Management Service Operations

Operate and troubleshoot NVIDIA DMS for centralized BlueField and ConnectX management.

Automation & Ops ✓ NVIDIA · Official

TAO NVIDIA GPU Host Setup

Checks and standardizes NVIDIA drivers, CUDA, and container runtime prerequisites for TAO GPU hosts.

Automation & Ops ✓ NVIDIA · Official

Dynamo Router Starter

Configure Dynamo routing modes and verify frontend endpoints quickly.

Automation & Ops ✓ NVIDIA · Official

Dynamo Interconnect Readiness Check

Verify RDMA, NVLink, and NIXL transport readiness before trusting performance results from disaggregated Dynamo serving.

Automation & Ops ✓ NVIDIA · Official

Dynamo Deployment Troubleshooter

Diagnose unhealthy Dynamo deployments on Kubernetes from read-only evidence.

Automation & Ops ✓ NVIDIA · Official

Dynamo Recipe Runner

Select, validate, minimally patch, and deploy existing NVIDIA Dynamo inference recipes on Kubernetes.

Automation & Ops ✓ NVIDIA · Official

Physical AI Infrastructure Resilience

Set up, validate, scale, and recover Physical AI infrastructure for synthetic data generation workflows.

Automation & Ops ✓ NVIDIA · Official

TAO Inference Microservice Deployment

Start, query, and stop TAO inference microservices across supported compute platforms.

Automation & Ops ✓ NVIDIA · Official

NVIDIA RAG Blueprint Operations

Deploy, configure, troubleshoot, and manage NVIDIA RAG Blueprint environments.

Dev & Engineering ✓ NVIDIA · Official

NeMo AutoModel Distributed Training

Choose and configure multi-GPU and multi-node training strategies for NeMo AutoModel.

Dev & Engineering ✓ NVIDIA · Official

NeMo AutoModel Recipe Development

Build, modify, and validate NeMo AutoModel training and evaluation recipes.

Automation & Ops ✓ NVIDIA · Official

NeMo AutoModel Launcher Configuration

Configure NeMo AutoModel launches across local, Slurm, and cloud runs.

Dev & Engineering ✓ NVIDIA · Official

MoE Hardware Configuration Reference

Plan MoE training layouts and throughput expectations across NVIDIA GPU platforms.

Dev & Engineering ✓ NVIDIA · Official

NeMo Relay Installation Guide

Choose and verify the right NeMo Relay path for CLIs, language packages, and maintained frameworks.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge GPU Memory Tuning

Diagnose GPU OOMs in Megatron Bridge training and reduce fragmentation, activation, and PEFT memory usage.

Related skills