kubeflow-trainer

kubeflow-trainer is a skill for Claude Code, Codex from Aidas-dev/k8s-agent-skills. It costs 56 tokens per session (3,236 once invoked), scanned A, original, MIT.

A guide for Kubeflow Trainer, a Kubernetes system for running machine-learning training jobs. It uses shared job definitions for frameworks such as PyTorch, TensorFlow, JAX, XGBoost, MPI, and Flux.

In plain words
What is it for?
Use it to define single- or multi-node training jobs, create reusable training runtimes, and run framework-specific workloads on Kubernetes.
Why use it?
It gives distributed training workloads a common Kubernetes configuration instead of requiring separate setup for each machine-learning framework.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to define single- or multi-node training jobs, create reusable training runtimes, and run framework-specific workloads on Kubernetes.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/aidas-dev/k8s-agent-skills/kubeflow-trainer
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add Aidas-dev/k8s-agent-skills --skill kubeflow-trainer
Clone the repo
git clone --depth 1 https://github.com/Aidas-dev/k8s-agent-skills

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for kubeflow-trainer

README.md
[![agentmods](https://agentmods.dev/badge/skills/aidas-dev/k8s-agent-skills/kubeflow-trainer/github.svg)](https://agentmods.dev/skills/aidas-dev/k8s-agent-skills/kubeflow-trainer)
Your own site
<a href="https://agentmods.dev/skills/aidas-dev/k8s-agent-skills/kubeflow-trainer"><img src="https://agentmods.dev/badge/skills/aidas-dev/k8s-agent-skills/kubeflow-trainer/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for kubeflow-trainer

Your own site · 80×15
<a href="https://agentmods.dev/skills/aidas-dev/k8s-agent-skills/kubeflow-trainer"><img src="https://agentmods.dev/badge/skills/aidas-dev/k8s-agent-skills/kubeflow-trainer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 56 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,236 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00056 $0.03236
Opus 5 $0.00028 $0.01618
Sonnet 5 $0.00011 $0.00647
Haiku 4.5 $0.00006 $0.00324

Measured 9d ago against content hash 519749ab1d07, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

kubeflow-trainer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/kubeflow-trainer/SKILL.md · 430 lines

How it starts

The opening of the file, as written. The whole thing — 430 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Kubeflow Trainer v2.2

Repository: github.com/kubeflow/trainer
Latest release: v2.2.0 (March 20, 2026)
API version: trainer.kubeflow.org/v1alpha1
Kubernetes: v1.36
Underlying orchestration: JobSet v0.10.1

Architecture

Single TrainJob CRD replaces framework-specific CRDs. TrainingRuntime defines the execution template. Framework-agnostic: one abstraction for PyTorch, TensorFlow, JAX, XGBoost, MPI, Flux.

TrainJob (training intent)
  └─ runtimeRef ──> TrainingRuntime / ClusterTrainingRuntime (execution template)
                     └─ template ──> JobSet (underlying orchestration)
                                      └─ replicatedJobs ──> Jobs ──> Pods

CRDs

TrainJob (namespaced)

Defines the training workload.

apiVersion: trainer.kubeflow.org/v1alpha1
kind: TrainJob
metadata:
  name: pytorch-mnist
spec:
  runtimeRef:
    name: pytorch-mnist-runtime
    apiGroup: trainer.kubeflow.org

  # Trainer definition
  trainer:
    image: pytorch/pytorch:2.5.0-cuda12.4-cudnn9-runtime
    command: ["python", "/workspace/train.py"]
    numNodes: 4
    env:
      - name: EPOCHS
        value: "10"
      - name: BATCH_SIZE
        value: "128"
    resources:
      requests:
        cpu: 4
        memory: 16Gi
      limits:
        nvidia.com/gpu: 2

  # Optional: dataset/model initializer (sidecar that runs before training)
  initializer:
    storageUri: s3://bucket/datasets/mnist/
    env:
      - name: AWS_ENDPOINT_URL
        value: s3.example.com
    secretRef:
      name: s3-credentials

  # Runtime patches (replaces deprecated PodTemplateOverrides)
  runtimePatches:
    - managerKey: user-overrides
      patch:
        - op: add
          path: /spec/replicatedJobs/0/template/spec/template/spec/containers/0/env/-
          value:
            name: LOG_LEVEL
            value: debug

  # Hard deadline
  activeDeadlineSeconds: 3600

  # Kueue integration
  managedBy: kueue.x-k8s.io/multikueue
Spec Fields

Read the full file on GitHub · 430 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 430 lines · 56 tokens per session scan A 519749ab1d07

Subscribe to this mod's changes

kubeflow-trainer is a skill published in the GitHub repository Aidas-dev/k8s-agent-skills (2 stars, last pushed 26d ago), licensed MIT. It adds 56 tokens to every session and 3,236 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

tensorrt-llm

High-throughput LLM inference on NVIDIA GPUs.

NousResearch/hermes-agent · 18 tokens

google-cloud-solution-guided-gke-ai-migration

Guides the migration of existing AI workloads (Cloud Run, Gemini API, Gemini Enterprise Agent Platform) to self-hosted GKE inference using gcloud and kubectl. Use when the user has an existing AI inference workload (on Cloud Run, the Gemini API, Gemini Enterprise Agent Platform, or a custom VM) and wants to move it to…

google/skills · 157 tokens

agent-platform-tuning

Agent Platform Model Tuning. Use when you need to fine-tune open models or Gemini models using Agent Platform infrastructure. Don't use for model training outside Agent Platform, model deployment to endpoints (use agent-platform-deploy), or managing serving endpoints (use agent-platform-endpoint-management).

google/skills · 64 tokens

modal

Modal is a serverless cloud platform for running Python on demand, including on-demand GPUs. Use when deploying or serving AI/ML models, running GPU-accelerated workloads (training, fine-tuning, inference), serving web endpoints, scheduling batch jobs, or scaling Python code to cloud containers with the Modal SDK.

K-Dense-AI/scientific-agent-skills · 65 tokens

gke-inference

Deploys and optimizes AI/ML inference workloads on GKE, using GPUs, TPUs, and model servers. Use when deploying GKE inference servers, configuring GKE GPU resources for inference, or deploying LLMs on GKE. Don't use for generic batch jobs or HPC task queues (use gke-batch-hpc instead).

google/skills · 74 tokens

agent-platform-endpoint-management

Manages Agent Platform serving endpoints. Use when you need to create, list, describe, update, or delete serving endpoints for model deployment on Agent Platform. Also use when troubleshooting endpoint permission, quota, or resource busy errors. Don't use for deploying models to endpoints or for running model…

google/skills · 64 tokens