[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7887X38XJEZWF79FVB7W0GS":3,"skill-same:01M7887X38XJEZWF79FVB7W0GS":59},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":26,"commitSha":27,"skillDirectory":28,"forksCount":14,"repositoryHtmlUrl":29,"repositoryHomepageUrl":30,"ownerHtmlUrl":31,"ownerBlogUrl":32,"ownerDescription":33,"updatedAt":34,"install":35,"resources":48},"01M7887X38XJEZWF79FVB7W0GS","gke-batch-hpc",null,"Runs batch and HPC workloads on GKE, utilizing job queues and parallel processing. Use when running GKE batch jobs, configuring GKE HPC, or setting up GKE job queues. Don't use for standard web application deployments (use gke-app-onboarding instead).","Solizardking\u002Fskillhub-main","Solizardking","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F108096810?v=4","01M786KDB10AYV042N0F9BD5B5","Clawd","01M786NYYXSZK1D05WM8JQPC48",0,"MIT","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fpublic\u002Fapi\u002Fskills\u002Fgoogle\u002Fcloud\u002Fgke-batch-hpc\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: gke-batch-hpc\ndescription: >-\n  Runs batch and HPC workloads on GKE, utilizing job queues and parallel\n  processing. Use when running GKE batch jobs, configuring GKE HPC, or setting\n  up GKE job queues. Don't use for standard web application deployments (use\n  gke-app-onboarding instead).\nmetadata:\n  category: Containers\n---\n\n# GKE Batch & HPC Workloads\n\nThis reference covers running batch processing and high-performance computing\n(HPC) workloads on GKE.\n\n> **MCP Tools:** `apply_k8s_manifest`, `get_k8s_resource`,\n> `describe_k8s_resource`, `get_k8s_logs`, `delete_k8s_resource`,\n> `list_k8s_events`\n\n## When to Use\n\n-   Running batch data processing pipelines\n-   HPC simulations (CFD, molecular dynamics, financial modeling)\n-   Large-scale parallel computation (MPI, MapReduce)\n-   ML training jobs\n-   CI\u002FCD build farms\n\n## Batch Processing on GKE\n\n### Kubernetes Jobs\n\n```yaml\napiVersion: batch\u002Fv1\nkind: Job\nmetadata:\n  name: batch-job\nspec:\n  parallelism: 10\n  completions: 100\n  backoffLimit: 3\n  template:\n    spec:\n      containers:\n      - name: worker\n        image: \u003CIMAGE>\n        resources:\n          requests:\n            cpu: \"1\"\n            memory: \"2Gi\"\n      restartPolicy: Never\n```\n\n### JobSet (for Complex Multi-Job Workflows)\n\nThe golden path enables JobSet monitoring (`JOBSET` in monitoringConfig).\n\n```yaml\napiVersion: jobset.x-k8s.io\u002Fv1alpha2\nkind: JobSet\nmetadata:\n  name: training-job\nspec:\n  replicatedJobs:\n  - name: workers\n    replicas: 4\n    template:\n      spec:\n        parallelism: 1\n        completions: 1\n        template:\n          spec:\n            containers:\n            - name: worker\n              image: \u003CIMAGE>\n              resources:\n                requests:\n                  cpu: \"4\"\n                  memory: \"8Gi\"\n```\n\n### Kueue (Job Queuing)\n\nKueue manages job scheduling and resource allocation for batch workloads:\n\n```bash\n# Install Kueue\nkubectl apply --server-side -f https:\u002F\u002Fgithub.com\u002Fkubernetes-sigs\u002Fkueue\u002Freleases\u002Flatest\u002Fdownload\u002Fmanifests.yaml\n```\n\n```yaml\n# Define a ClusterQueue\napiVersion: kueue.x-k8s.io\u002Fv1beta1\nkind: ClusterQueue\nmetadata:\n  name: batch-queue\nspec:\n  namespaceSelector: {}\n  resourceGroups:\n  - coveredResources: [\"cpu\", \"memory\"]\n    flavors:\n    - name: default\n      resources:\n      - name: \"cpu\"\n        nominalQuota: 100\n      - name: \"memory\"\n        nominalQuota: \"200Gi\"\n---\n# Allow a namespace to use the queue\napiVersion: kueue.x-k8s.io\u002Fv1beta1\nkind: LocalQueue\nmetadata:\n  name: batch-local\n  namespace: batch-jobs\nspec:\n  clusterQueue: batch-queue\n```\n\n## HPC on GKE\n\n### Compact Placement (Low-Latency Networking)\n\nFor tightly-coupled HPC workloads that need low-latency inter-node\ncommunication:\n\n```bash\n# Standard clusters: create node pool with compact placement\ngcloud container node-pools create hpc-pool \\\n  --cluster \u003CCLUSTER_NAME> --region \u003CREGION> \\\n  --machine-type c3-standard-44 \\\n  --placement-type COMPACT \\\n  --num-nodes 8 \\\n  --enable-autoscaling --min-nodes 0 --max-nodes 16 \\\n  --quiet\n```\n\n### MPI Workloads\n\nUse the MPI Operator for MPI-based HPC applications:\n\n```bash\n# Install MPI Operator\nkubectl apply -f https:\u002F\u002Fraw.githubusercontent.com\u002Fkubeflow\u002Fmpi-operator\u002Fmaster\u002Fdeploy\u002Fv2beta1\u002Fmpi-operator.yaml\n```\n\n```yaml\napiVersion: kubeflow.org\u002Fv2beta1\nkind: MPIJob\nmetadata:\n  name: hpc-simulation\nspec:\n  slotsPerWorker: 4\n  mpiReplicaSpecs:\n    Launcher:\n      replicas: 1\n      template:\n        spec:\n          containers:\n          - name: launcher\n            image: \u003CMPI_IMAGE>\n            command: [\"mpirun\", \"-np\", \"32\", \".\u002Fsimulation\"]\n            resources:\n              requests:\n                cpu: \"1\"\n                memory: \"2Gi\"\n              limits:\n                cpu: \"2\"\n                memory: \"4Gi\"\n    Worker:\n      replicas: 8\n      template:\n        spec:\n          containers:\n          - name: worker\n            image: \u003CMPI_IMAGE>\n            resources:\n              requests:\n                cpu: \"4\"\n                memory: \"8Gi\"\n              limits:\n                cpu: \"8\"\n                memory: \"16Gi\"\n```\n\n## Cost Optimization for Batch\u002FHPC\n\n### Spot VMs for Batch\n\nBatch workloads are ideal Spot VM candidates (interruptible, can checkpoint).\nUse a ComputeClass with Spot-first priority and `activeMigration` to return to\nSpot when available. See the `gke-compute-classes` skill for the\nSpot-with-fallback pattern.\n\n### Scale-to-Zero\n\nFor batch clusters, allow node pools to scale to zero when no jobs are running:\n\n-   Autopilot (golden path): Automatic, nodes scale to zero when no pods are\n    scheduled\n-   Standard: Set `--min-nodes 0` on batch node pools\n\n## Best Practices & Production Guidelines\n\n-   **Resource Quotas**: Always specify resource requests and limits (CPU,\n    memory, and optionally GPU\u002FTPU) for all batch\u002FHPC manifests. This is\n    critical for Kueue admission, autoscaling, and preventing resource\n    starvation in the cluster.\n-   **TPU\u002FSpot Cluster Maintenance**: For long-running AI training runs on Spot\n    VMs\u002FTPUs, advise using **GKE maintenance exclusions** to block automatic\n    cluster upgrades\u002Freboots during the active training window to minimize\n    unnecessary preemption.\n-   **MPI Workloads**: Use the **Kubeflow Training Operator** to orchestrate\n    distributed MPI applications via the `MPIJob` custom resource.\n-   **Kueue & JobSet**: Use **Kueue** for multi-tenant job queueing and fair\n    sharing; use **JobSet** for multi-component tightly coupled workloads.\n-   **Resilience**: Always set a `backoffLimit` on Jobs, and implement\n    application-level checkpointing (e.g., using Orbax or PyTorch checkpointing)\n    to survive Spot VM preemption.\n","# GKE Batch & HPC Workloads\n\nThis reference covers running batch processing and high-performance computing\n(HPC) workloads on GKE.\n\n> **MCP Tools:** `apply_k8s_manifest`, `get_k8s_resource`,\n> `describe_k8s_resource`, `get_k8s_logs`, `delete_k8s_resource`,\n> `list_k8s_events`\n\n## When to Use\n\n-   Running batch data processing pipelines\n-   HPC simulations (CFD, molecular dynamics, financial modeling)\n-   Large-scale parallel computation (MPI, MapReduce)\n-   ML training jobs\n-   CI\u002FCD build farms\n\n## Batch Processing on GKE\n\n### Kubernetes Jobs\n\n```yaml\napiVersion: batch\u002Fv1\nkind: Job\nmetadata:\n  name: batch-job\nspec:\n  parallelism: 10\n  completions: 100\n  backoffLimit: 3\n  template:\n    spec:\n      containers:\n      - name: worker\n        image: \u003CIMAGE>\n        resources:\n          requests:\n            cpu: \"1\"\n            memory: \"2Gi\"\n      restartPolicy: Never\n```\n\n### JobSet (for Complex Multi-Job Workflows)\n\nThe golden path enables JobSet monitoring (`JOBSET` in monitoringConfig).\n\n```yaml\napiVersion: jobset.x-k8s.io\u002Fv1alpha2\nkind: JobSet\nmetadata:\n  name: training-job\nspec:\n  replicatedJobs:\n  - name: workers\n    replicas: 4\n    template:\n      spec:\n        parallelism: 1\n        completions: 1\n        template:\n          spec:\n            containers:\n            - name: worker\n              image: \u003CIMAGE>\n              resources:\n                requests:\n                  cpu: \"4\"\n                  memory: \"8Gi\"\n```\n\n### Kueue (Job Queuing)\n\nKueue manages job scheduling and resource allocation for batch workloads:\n\n```bash\n# Install Kueue\nkubectl apply --server-side -f https:\u002F\u002Fgithub.com\u002Fkubernetes-sigs\u002Fkueue\u002Freleases\u002Flatest\u002Fdownload\u002Fmanifests.yaml\n```\n\n```yaml\n# Define a ClusterQueue\napiVersion: kueue.x-k8s.io\u002Fv1beta1\nkind: ClusterQueue\nmetadata:\n  name: batch-queue\nspec:\n  namespaceSelector: {}\n  resourceGroups:\n  - coveredResources: [\"cpu\", \"memory\"]\n    flavors:\n    - name: default\n      resources:\n      - name: \"cpu\"\n        nominalQuota: 100\n      - name: \"memory\"\n        nominalQuota: \"200Gi\"\n---\n# Allow a namespace to use the queue\napiVersion: kueue.x-k8s.io\u002Fv1beta1\nkind: LocalQueue\nmetadata:\n  name: batch-local\n  namespace: batch-jobs\nspec:\n  clusterQueue: batch-queue\n```\n\n## HPC on GKE\n\n### Compact Placement (Low-Latency Networking)\n\nFor tightly-coupled HPC workloads that need low-latency inter-node\ncommunication:\n\n```bash\n# Standard clusters: create node pool with compact placement\ngcloud container node-pools create hpc-pool \\\n  --cluster \u003CCLUSTER_NAME> --region \u003CREGION> \\\n  --machine-type c3-standard-44 \\\n  --placement-type COMPACT \\\n  --num-nodes 8 \\\n  --enable-autoscaling --min-nodes 0 --max-nodes 16 \\\n  --quiet\n```\n\n### MPI Workloads\n\nUse the MPI Operator for MPI-based HPC applications:\n\n```bash\n# Install MPI Operator\nkubectl apply -f https:\u002F\u002Fraw.githubusercontent.com\u002Fkubeflow\u002Fmpi-operator\u002Fmaster\u002Fdeploy\u002Fv2beta1\u002Fmpi-operator.yaml\n```\n\n```yaml\napiVersion: kubeflow.org\u002Fv2beta1\nkind: MPIJob\nmetadata:\n  name: hpc-simulation\nspec:\n  slotsPerWorker: 4\n  mpiReplicaSpecs:\n    Launcher:\n      replicas: 1\n      template:\n        spec:\n          containers:\n          - name: launcher\n            image: \u003CMPI_IMAGE>\n            command: [\"mpirun\", \"-np\", \"32\", \".\u002Fsimulation\"]\n            resources:\n              requests:\n                cpu: \"1\"\n                memory: \"2Gi\"\n              limits:\n                cpu: \"2\"\n                memory: \"4Gi\"\n    Worker:\n      replicas: 8\n      template:\n        spec:\n          containers:\n          - name: worker\n            image: \u003CMPI_IMAGE>\n            resources:\n              requests:\n                cpu: \"4\"\n                memory: \"8Gi\"\n              limits:\n                cpu: \"8\"\n                memory: \"16Gi\"\n```\n\n## Cost Optimization for Batch\u002FHPC\n\n### Spot VMs for Batch\n\nBatch workloads are ideal Spot VM candidates (interruptible, can checkpoint).\nUse a ComputeClass with Spot-first priority and `activeMigration` to return to\nSpot when available. See the `gke-compute-classes` skill for the\nSpot-with-fallback pattern.\n\n### Scale-to-Zero\n\nFor batch clusters, allow node pools to scale to zero when no jobs are running:\n\n-   Autopilot (golden path): Automatic, nodes scale to zero when no pods are\n    scheduled\n-   Standard: Set `--min-nodes 0` on batch node pools\n\n## Best Practices & Production Guidelines\n\n-   **Resource Quotas**: Always specify resource requests and limits (CPU,\n    memory, and optionally GPU\u002FTPU) for all batch\u002FHPC manifests. This is\n    critical for Kueue admission, autoscaling, and preventing resource\n    starvation in the cluster.\n-   **TPU\u002FSpot Cluster Maintenance**: For long-running AI training runs on Spot\n    VMs\u002FTPUs, advise using **GKE maintenance exclusions** to block automatic\n    cluster upgrades\u002Freboots during the active training window to minimize\n    unnecessary preemption.\n-   **MPI Workloads**: Use the **Kubeflow Training Operator** to orchestrate\n    distributed MPI applications via the `MPIJob` custom resource.\n-   **Kueue & JobSet**: Use **Kueue** for multi-tenant job queueing and fair\n    sharing; use **JobSet** for multi-component tightly coupled workloads.\n-   **Resilience**: Always set a `backoffLimit` on Jobs, and implement\n    application-level checkpointing (e.g., using Orbax or PyTorch checkpointing)\n    to survive Spot VM preemption.\n",{"name":5,"metadata":24,"description":7},{"category":25},"Containers","main","be3de0244dc1a57e323d8160787ebb40fb64930e","public\u002Fapi\u002Fskills\u002Fgoogle\u002Fcloud\u002Fgke-batch-hpc","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main","https:\u002F\u002Fskills.x402.wtf","https:\u002F\u002Fgithub.com\u002FSolizardking","https:\u002F\u002Fmusebook.trade","Got a mile high and never came down","2026-10-08 23:20:13",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add Solizardking\u002Fskillhub-main --skill gke-batch-hpc",{"tool":41,"command":42},"bunx","bunx skills add Solizardking\u002Fskillhub-main --skill gke-batch-hpc",{"tool":44,"command":45},"pnpm","pnpm dlx skills add Solizardking\u002Fskillhub-main --skill gke-batch-hpc","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fpublic\u002Fapi\u002Fskills\u002Fgoogle\u002Fcloud\u002Fgke-batch-hpc\u002FSKILL.md",true,[49,55],{"path":50,"resourceType":51,"resourceTypeDesc":52,"sizeBytes":53,"sourceHtmlUrl":54},"metadata.json","REFERENCE","参考文档",436,"https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fpublic\u002Fapi\u002Fskills\u002Fgoogle\u002Fcloud\u002Fgke-batch-hpc\u002Fmetadata.json",{"path":56,"resourceType":51,"resourceTypeDesc":52,"sizeBytes":57,"sourceHtmlUrl":58},"verification.json",1130,"https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fpublic\u002Fapi\u002Fskills\u002Fgoogle\u002Fcloud\u002Fgke-batch-hpc\u002Fverification.json",[60,67,74,81,88,96],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":19},"01M78E0TGCAMNPS207BA3G07S7","zkrouter","Self-hosted OpenAI-compatible LLM router — smart tier-based routing, ZK-stamped routing decisions, mode overrides, free OpenRouter at install via the Birth bot. Save 60-80% on inference costs by routing to the cheapest capable model across 12+ OpenRouter models.","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fzkrouter\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":19},"01M78E0R1Z8BDT6EDTHY17MTNB","zk","For custom ZK Solana programs and privacy-preserving applications to prevent double spending. Guide to integrate rent-free nullifier PDAs for double-spend prevention.","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fzk\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":19},"01M78E0P3ZTZ1YWF8MD5MM23QE","zk-omni-messaging","Zero-knowledge omnichain messaging between Robinhood Chain and Solana using CheshireZkOmniMessenger (msgType 4) with nullifier anti-replay and the zk-omni-relayer service.","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fzk-omni-messaging\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":19},"01M78E0JXZ2PNB1896VXAGDBWD","zerion-wallet-data","Use when the user asks to check a wallet's portfolio, token balances, DeFi positions, PnL, transaction history, or NFTs. Provides interpreted on-chain data across 40+ EVM chains and Solana via x402 pay-per-request or API key — no raw RPC decoding needed.","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fzerion-wallet-data\u002FSKILL.md",[],[],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":92,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M78E0G14FH9GRRTQYMKD5E59","youtube-clipper","YouTube 视频智能剪辑工具。下载视频和字幕，AI 分析生成精细章节（几分钟级别）， 用户选择片段后自动剪辑、翻译字幕为中英双语、烧录字幕到视频，并生成总结文案。 使用场景：当用户需要剪辑 YouTube 视频、生成短视频片段、制作双语字幕版本时。 关键词：视频剪辑、YouTube、字幕翻译、双语字幕、视频下载、clip video","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fyoutube-clipper\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fyoutube-clipper\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M78E0C2ZYA99CFFDNB7ZMK5S","yield-optimization","Multi-chain yield optimization via yield.xyz (StakeKit). Discover 2,988+ yield opportunities across 75+ blockchains, build deposit\u002Fwithdrawal transactions with shell scripts, and sign them with a MoonPay wallet. Use when the user wants to earn yield, find the best lending or staking rate, enter\u002Fexit a position, or check portfolio balances.","https:\u002F\u002Fgithub.com\u002FSolizardking\u002Fskillhub-main\u002Fblob\u002Fmain\u002Fskills\u002Fyield-optimization\u002FSKILL.md",[],[]]