[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skills:stats":3,"skills:pro6":8,"skills:sce6":55,"skills:creators9":100,"skills:all:stars:248:":195},{"skillCount":4,"professionCount":5,"scenarioCount":6,"creatorCount":7},463411,11,7,5744,[9,20,27,34,41,48],{"categoryNo":10,"taxonomyType":11,"taxonomyTypeDesc":12,"categoryName":13,"categoryNameEn":14,"description":15,"level":16,"skillCount":17,"parentNo":18,"parentName":18,"children":19},"01JBSEEDCAT000000000000030","PROFESSION","适用职业","开发工程师","Software Engineer","从事软件编码、测试与运维的工程师",1,13341,null,[],{"categoryNo":21,"taxonomyType":11,"taxonomyTypeDesc":12,"categoryName":22,"categoryNameEn":23,"description":24,"level":16,"skillCount":25,"parentNo":18,"parentName":18,"children":26},"01JBSEEDCAT000000000000032","设计师","Designer","视觉、交互与品牌设计人员",7902,[],{"categoryNo":28,"taxonomyType":11,"taxonomyTypeDesc":12,"categoryName":29,"categoryNameEn":30,"description":31,"level":16,"skillCount":32,"parentNo":18,"parentName":18,"children":33},"01JBSEEDCAT000000000000033","数据分析师","Data Analyst","数据分析、报表与商业智能人员",5005,[],{"categoryNo":35,"taxonomyType":11,"taxonomyTypeDesc":12,"categoryName":36,"categoryNameEn":37,"description":38,"level":16,"skillCount":39,"parentNo":18,"parentName":18,"children":40},"01JBSEEDCAT000000000000036","内容创作者","Content Creator","作者、编辑与自媒体创作者",3978,[],{"categoryNo":42,"taxonomyType":11,"taxonomyTypeDesc":12,"categoryName":43,"categoryNameEn":44,"description":45,"level":16,"skillCount":46,"parentNo":18,"parentName":18,"children":47},"01JBSEEDCAT000000000000035","市场营销","Marketer","市场推广、增长与品牌运营人员",3454,[],{"categoryNo":49,"taxonomyType":11,"taxonomyTypeDesc":12,"categoryName":50,"categoryNameEn":51,"description":52,"level":16,"skillCount":53,"parentNo":18,"parentName":18,"children":54},"01JBSEEDCAT000000000000037","律师与法务","Legal","合同、合规与法律事务人员",3337,[],[56,65,72,79,86,93],{"categoryNo":57,"taxonomyType":58,"taxonomyTypeDesc":59,"categoryName":60,"categoryNameEn":61,"description":62,"level":16,"skillCount":63,"parentNo":18,"parentName":18,"children":64},"01JBSEEDCAT000000000000022","SCENARIO","使用场景","智能体与自动化","Agents & Automation","Skill、MCP、智能体与工作流自动化",39675,[],{"categoryNo":66,"taxonomyType":58,"taxonomyTypeDesc":59,"categoryName":67,"categoryNameEn":68,"description":69,"level":16,"skillCount":70,"parentNo":18,"parentName":18,"children":71},"01JBSEEDCAT000000000000006","软件开发","Software Development","编码、测试、部署等软件研发任务",26798,[],{"categoryNo":73,"taxonomyType":58,"taxonomyTypeDesc":59,"categoryName":74,"categoryNameEn":75,"description":76,"level":16,"skillCount":77,"parentNo":18,"parentName":18,"children":78},"01JBSEEDCAT000000000000014","设计创意","Design & Creative","视觉设计、品牌与多媒体创作",13847,[],{"categoryNo":80,"taxonomyType":58,"taxonomyTypeDesc":59,"categoryName":81,"categoryNameEn":82,"description":83,"level":16,"skillCount":84,"parentNo":18,"parentName":18,"children":85},"01JBSEEDCAT000000000000001","办公效率","Office Productivity","日常办公文档、表格、演示与沟通类任务",6403,[],{"categoryNo":87,"taxonomyType":58,"taxonomyTypeDesc":59,"categoryName":88,"categoryNameEn":89,"description":90,"level":16,"skillCount":91,"parentNo":18,"parentName":18,"children":92},"01JBSEEDCAT000000000000017","内容创作","Content Creation","写作、文案与营销内容生产",6063,[],{"categoryNo":94,"taxonomyType":58,"taxonomyTypeDesc":59,"categoryName":95,"categoryNameEn":96,"description":97,"level":16,"skillCount":98,"parentNo":18,"parentName":18,"children":99},"01JBSEEDCAT000000000000020","研究与学习","Research & Learning","资料检索、知识整理与学习辅导",5574,[],{"records":101,"total":7,"current":16,"size":194},[102,114,123,132,142,152,162,172,182],{"accountNo":103,"login":104,"displayName":105,"avatarUrl":106,"accountType":107,"accountTypeDesc":108,"profileDescription":109,"skillCount":110,"repoCount":111,"professionCount":5,"scenarioCount":6,"totalStars":112,"updatedAt":113},"01M698VYDXZ6JW5WH3CRV650R8","microsoft","Microsoft","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6154722?v=4","ORGANIZATION","组织","Open source projects and samples from Microsoft",4300,236,832929,"2026-11-15 04:18:35",{"accountNo":115,"login":116,"displayName":116,"avatarUrl":117,"accountType":107,"accountTypeDesc":108,"profileDescription":118,"skillCount":119,"repoCount":120,"professionCount":6,"scenarioCount":6,"totalStars":121,"updatedAt":122},"01M698P3FSAZ6B4R928ZF19KPT","openclaw","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F252820863?v=4","Your personal, open source AI assistant.",369,32,437811,"2026-11-15 03:04:02",{"accountNo":124,"login":125,"displayName":126,"avatarUrl":127,"accountType":107,"accountTypeDesc":108,"profileDescription":18,"skillCount":128,"repoCount":129,"professionCount":5,"scenarioCount":6,"totalStars":130,"updatedAt":131},"01M697XGDPXJAVQ8391V2BCYTE","anthropics","Anthropic","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F76263028?v=4",716,18,427323,"2026-11-15 05:52:49",{"accountNo":133,"login":134,"displayName":135,"avatarUrl":136,"accountType":107,"accountTypeDesc":108,"profileDescription":18,"skillCount":137,"repoCount":138,"professionCount":6,"scenarioCount":139,"totalStars":140,"updatedAt":141},"01M6AC29YGCR4S2NTJFV0G8B81","openai","OpenAI","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F14957082?v=4",177,23,6,344400,"2026-11-15 03:07:28",{"accountNo":143,"login":144,"displayName":145,"avatarUrl":146,"accountType":107,"accountTypeDesc":108,"profileDescription":18,"skillCount":147,"repoCount":148,"professionCount":149,"scenarioCount":6,"totalStars":150,"updatedAt":151},"01M6AC83E6K4FDKK7HZ8WNMQKE","apache","The Apache Software Foundation","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F47359?v=4",440,76,8,298252,"2026-11-15 04:11:16",{"accountNo":153,"login":154,"displayName":155,"avatarUrl":156,"accountType":107,"accountTypeDesc":108,"profileDescription":157,"skillCount":158,"repoCount":159,"professionCount":5,"scenarioCount":6,"totalStars":160,"updatedAt":161},"01M6AC83DS133TPXFSP1DHNCC6","github","GitHub","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F9919?v=4","How people build software.",1022,30,278314,"2026-11-15 06:45:15",{"accountNo":163,"login":164,"displayName":165,"avatarUrl":166,"accountType":107,"accountTypeDesc":108,"profileDescription":167,"skillCount":168,"repoCount":169,"professionCount":5,"scenarioCount":6,"totalStars":170,"updatedAt":171},"01M698VYEFDM0VJBZDYYJ2T83S","google","Google","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1342004?v=4","Google ❤️ Open Source",542,50,243804,"2026-11-15 03:00:30",{"accountNo":173,"login":174,"displayName":175,"avatarUrl":176,"accountType":107,"accountTypeDesc":108,"profileDescription":18,"skillCount":177,"repoCount":178,"professionCount":179,"scenarioCount":6,"totalStars":180,"updatedAt":181},"01M6AC83EPDGBFN19MWYRREJD0","Significant-Gravitas","AutoGPT","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F130738209?v=4",87,2,10,188029,"2026-11-15 06:21:09",{"accountNo":183,"login":184,"displayName":185,"avatarUrl":186,"accountType":187,"accountTypeDesc":188,"profileDescription":189,"skillCount":190,"repoCount":191,"professionCount":139,"scenarioCount":6,"totalStars":192,"updatedAt":193},"01M6DJCCDDG53MTYYRTXAXFG17","garrytan","Garry Tan","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F19957?v=4","USER","个人作者","Writes software, dreams. ",340,3,175019,"2026-11-15 03:57:03",9,{"records":196,"total":4,"current":518,"size":129},[197,229,251,272,291,310,339,363,382,397,406,423,436,451,465,476,492,505],{"skillNo":198,"name":199,"nameZh":18,"description":200,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":205,"categories":206,"keywords":207,"publishedAt":228},"01M6993QSTRR35YRH6F5HGKJT5","gke-cluster-autoscaler","Trigger on mention of GKE cluster autoscaler, node autoscaling, node pool auto-creation \u002F node auto-provisioning. Provides guidance on enabling and optimizing cluster autoscaler, best practices, and troubleshooting issues such as nodes not scaling up or down, zonal stockouts, or capacity buffers. Do not use for ComputeClass-specific YAML generation or priority configuration (defer to gke-compute-classes skill).","google\u002Fskills","01M698WR0H76ENHE587T9EJT89",24235,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-cluster-autoscaler\u002FSKILL.md",[],[208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227],"gke cluster autoscaler","gke","cluster","cluster autoscaler","containers","capacity","best practices","issues","use","auto","up","do","compute","scaling","creation","as","trigger","autoscaling","for","down","2026-11-10 10:12:30",{"skillNo":230,"name":231,"nameZh":18,"description":232,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":233,"categories":234,"keywords":235,"publishedAt":228},"01M6993NZAP78DGV2RGK35R6AX","gke-batch-hpc","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).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-batch-hpc\u002FSKILL.md",[],[236,237,209,238,212,239,240,216,218,241,242,243,244,245,246,226,247,248,249,250],"gke batch hpc","batch","hpc","web","parallel","deployments","jobs","application","parallel processing","app","processing","queues","standard","when","running",{"skillNo":252,"name":253,"nameZh":18,"description":254,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":255,"categories":256,"keywords":257,"publishedAt":228},"01M6993KRQA9Y9QTTPH1AWXX9N","gke-basics","Manages core GKE cluster provisioning, credentials, Autopilot vs Standard selection, and workload deployment. Use when creating GKE clusters, fetching kubectl credentials, or deciding between Autopilot and Standard modes. Don't use for Workload Identity (use gke-workload-identity), GKE networking (use gke-networking), security hardening (use gke-platform-security or gke-workload-security), or cluster upgrades (use gke-upgrades).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-basics\u002FSKILL.md",[],[258,209,259,212,260,261,262,263,264,265,266,267,268,216,269,270,210,226,271,248,249],"gke basics","basics","security","gke networking","workload","deployment","core","platform-security","hardening","credentials","upgrades","selection","platform","modes",{"skillNo":273,"name":274,"nameZh":18,"description":275,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":276,"categories":277,"keywords":278,"publishedAt":228},"01M6993J99P7CKYMFR5P3CFF68","gke-backup-dr","Configures Backup for GKE: the BackupRestore cluster addon, BackupPlan and RestorePlan resources, restore workflows, and CMEK-encrypted backups. Use for backup policies, disaster recovery, or GKE cluster restores. Don't use for database backups.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-backup-dr\u002FSKILL.md",[],[279,209,280,281,282,283,284,285,216,286,287,288,210,289,290,226],"gke backup dr","backup","dr","storage","workflows","database","recovery","restore","disaster recovery","resources","encrypted","database backups",{"skillNo":292,"name":293,"nameZh":18,"description":294,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":295,"categories":296,"keywords":300,"publishedAt":228},"01M6993G44KHSR7KV34PYT4ACD","gke-app-onboarding","Manages GKE application onboarding, covering containerization, deployment manifests, and migration. Use when onboarding or deploying an application to GKE for the first time, or containerizing an app for GKE. Don't use for general GKE cluster administration or upgrades (use gke-basics or gke-upgrades instead).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-app-onboarding\u002FSKILL.md",[297],{"categoryNo":298,"categoryName":299,"taxonomyType":11,"taxonomyTypeDesc":12},"01JBSEEDCAT000000000000038","人力资源",[301,302,209,245,212,263,303,268,216,304,305,259,306,243,307,210,308,309,226,249],"gke app onboarding","onboarding","first","containerization","migration","general","time","manifests","administration",{"skillNo":311,"name":312,"nameZh":18,"description":313,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":314,"categories":315,"keywords":319,"publishedAt":228},"01M6993DSBN53K1A2NEJCK3VD2","gke-alert-configuration","Configures alerting policies in Terraform for Google Kubernetes Engine (GKE) clusters, workloads, and services using PromQL and Google Cloud Managed Service for Prometheus. Use when writing, analyzing, validating, or deploying Terraform alerting policies to monitor GKE service latency, traffic, error rates using Multi-Window Multi-Burn-Rate SLO alerts, memory saturation, and cluster health such as CrashLoopBackOff and Node NotReady conditions. Don't use for non-GKE compute runtimes such as standalone Compute Engine VMs or standalone Cloud Run services without GKE.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-alert-configuration\u002FSKILL.md",[316],{"categoryNo":317,"categoryName":318,"taxonomyType":58,"taxonomyTypeDesc":59},"01JBSEEDCAT000000000000011","运维与部署",[320,209,321,322,323,164,154,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,249],"gke alert configuration","alert","configuration","writing","kubernetes","google cloud","memory","alerting","slo","burn-rate","saturation","main","alerts","error","Managed Service","traffic","standalone","promql","managed",{"skillNo":340,"name":341,"nameZh":18,"description":342,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":343,"categories":344,"keywords":345,"publishedAt":228},"01M6993C2VAXF46BT2DRTNXCJ2","gke-ai-troubleshooting-tpu-vm-repair-stuck","Diagnose GKE TPU VM instances stuck in Compute Engine repair (REPAIRING state) after host errors, maintenance events, or faulty host reports in single-host, multi-host, dynamic slicing, and TPU All Capacity mode node pools. Use when TPU instances in a slice node pool remain in REPAIRING, host error events appear in audit logs, or TPU slices are unschedulable due to a VM in repair. Don't use for nodes reporting Ready: Unknown when no VM is in REPAIRING (use gke-ai-troubleshooting-node-unresponsive-timeout), other NotReady nodes when no VM is in REPAIRING (use gke-node-notready), Pending Pods when all nodes are Ready (use gke-workload-troubleshooting), JobSet preemption events (use gke-ai-troubleshooting-jobset-interruption), or Slice custom resource lifecycle failures (use gke-ai-troubleshooting-tpu-dynamic-slices-monitoring).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-vm-repair-stuck\u002FSKILL.md",[],[346,347,348,209,349,350,212,351,352,353,354,355,356,357,358,359,333,360,361,362,249],"gke ai troubleshooting tpu vm repair stuck","tpu","vm","troubleshooting","repair","audit","logs","monitoring","reporting","node","ready","reports","instances","host","errors","mode","unknown",{"skillNo":364,"name":365,"nameZh":18,"description":366,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":367,"categories":368,"keywords":372,"publishedAt":228},"01M6993AF0E61B2KXCBCFNC27P","gke-ai-troubleshooting-tpu-vbar-oom","Diagnoses and prevents vbar_control_agent segfaults, out-of-memory (OOM) errors, and TPU device initialization failures on TPU v6e nodes in GKE caused by race conditions during TPU device resets or high-frequency metrics polling. Use when troubleshooting vbar_control_agent crashes, memory cgroup OOMs in serial console logs, tpu-device-plugin metrics checksum corruption errors, or custom TPU metrics collection conflicts on GKE TPU v6e nodes. Don't use for general non-TPU container OOM troubleshooting or standard GKE node lifecycle operations.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-vbar-oom\u002FSKILL.md",[369],{"categoryNo":370,"categoryName":371,"taxonomyType":58,"taxonomyTypeDesc":59},"01JBSEEDCAT000000000000023","Skill与智能体开发",[373,347,209,349,374,326,375,352,376,377,355,216,378,306,379,360,226,380,381,249],"gke ai troubleshooting tpu vbar oom","oom","metrics","lifecycle","collection","container","conflicts","initialization","custom",{"skillNo":383,"name":384,"nameZh":18,"description":385,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":386,"categories":387,"keywords":388,"publishedAt":228},"01M699390WT2C0D71HN9395FCQ","gke-ai-troubleshooting-tpu-slice-mig-stuck-updating","Diagnose and resolve GKE TPU Dynamic Slice creation and JobSet scheduling failures blocked by underlying Compute Engine managed instance groups (MIGs) stuck in an updating or unstable state (isStable=False). Use when TPU slice creation fails with a SliceCreationFailed status reason, when JobSets remain stuck in pending, or when MIG status reflects isStable=False or versionTarget.isReached=False during TPU slice formation. Don't use for SliceCreationFailed causes unrelated to MIG stability (use gke-ai-troubleshooting-tpu-dynamic-slices-monitoring) or general pod scheduling failures unrelated to TPU slices.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-slice-mig-stuck-updating\u002FSKILL.md",[],[389,347,209,349,390,391,212,392,353,393,216,306,394,220,222,395,226,396,338,249],"gke ai troubleshooting tpu slice mig stuck updating","mig","slice","diagnose","engine","scheduling","false","pending",{"skillNo":398,"name":399,"nameZh":18,"description":400,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":401,"categories":402,"keywords":403,"publishedAt":228},"01M69937319KXABPP31MBSN31V","gke-ai-troubleshooting-tpu-reservation-stockout","Diagnose and resolve cluster-wide GKE TPU node pool creation or scale-up failures caused by Compute Engine reservation stockouts (GCE_STOCKOUT, ZONE_RESOURCE_POOL_EXHAUSTED) when existing unhealthy or NotReady TPU node pools in the cluster have node auto-repair disabled and hold reserved TPU capacity. Use when TPU node pool creation or scale-up fails with GCE_STOCKOUT, ZONE_RESOURCE_POOL_EXHAUSTED, or ZONE_RESOURCE_POOL_EXHAUSTED_WITH_DETAILS despite having a specific Compute Engine TPU reservation. Don't use for general pod scheduling without a TPU reservation or workload-level container crashes.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-reservation-stockout\u002FSKILL.md",[],[404,347,209,349,212,262,392,355,216,217,378,218,306,394,350,222,210,226,405,249],"gke ai troubleshooting tpu reservation stockout","Existing",{"skillNo":407,"name":408,"nameZh":18,"description":409,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":410,"categories":411,"keywords":412,"publishedAt":228},"01M699355JAT7QAZAKEQA3YARG","gke-ai-troubleshooting-tpu-performance-degradation","Diagnose GKE Cloud TPU training throughput drops and step-time regressions (15%+ TPU duty-cycle drop) using ML Diagnostics Workload Monitoring (`gcloud alpha mldiagnostics monitored-events` \u002F `hypercomputecluster.googleapis.com\u002Fv1alpha`) and 1-minute Cloud Monitoring system metrics (`kubernetes.io\u002Fnode\u002Faccelerator\u002F*`). Distinguishes hardware and network fabric throttling from workload resource bottlenecks (HBM capacity, host memory, or host CPU saturation). Use when TPU training throughput or duty cycle drops without crashing pods, when `PERFORMANCE_DEGRADATION` monitored events fire, or when triaging slow multi-slice training steps. Don't use for complete multi-slice XLA execution stalls with `HANG_DETECTED` logs (use gke-ai-troubleshooting-tpu-mxla-hang) or pod eviction\u002Finterruption restarts (use gke-ai-troubleshooting-jobset-interruption).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-performance-degradation\u002FSKILL.md",[],[413,414,347,209,349,415,416,324,326,417,418,352,353,355,330,419,420,307,359,421,422,249],"performance","gke ai troubleshooting tpu performance degradation","degradation","gcloud","training","alpha","cycle","drop","io","CPU",{"skillNo":424,"name":425,"nameZh":18,"description":426,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":427,"categories":428,"keywords":429,"publishedAt":228},"01M69933KTF7WXWFGKMGHPCZY0","gke-ai-troubleshooting-tpu-partition-overlap","Diagnose and resolve GKE TPU slice creation failures (SliceCreationFailed) caused by partition overlap conflicts where requested TPU partitions or their nodes are already used by another slice or JobSet. Use when diagnosing GKE TPU slice creation failures with SliceCreationFailed partition or node conflicts. Don't use for general slice lifecycle states or slice provisioning failures without node overlap (use gke-ai-troubleshooting-tpu-dynamic-slices-monitoring), quota exhaustion, or unresponsive TPU nodes.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-partition-overlap\u002FSKILL.md",[],[430,347,209,349,431,212,392,353,376,216,306,379,432,433,222,434,435,226,391,249],"gke ai troubleshooting tpu partition overlap","partition","quota","partitions","states","with",{"skillNo":437,"name":438,"nameZh":18,"description":439,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":440,"categories":441,"keywords":442,"publishedAt":228},"01M69931J2PCGJ8EDJ79BQPA76","gke-ai-troubleshooting-tpu-mxla-hang","Diagnose GKE Cloud TPU multi-slice training hangs (Megascale `HANG_DETECTED` logs and hang monitored events) using the ML Diagnostics `Megascale XLA (MXLA) Hang Analyzer` (`gcloud alpha mldiagnostics monitored-events`) and 1-minute Cloud Monitoring multi-slice latency metrics (`kubernetes.io\u002Fcontainer\u002Fmultislice\u002F*`). Distinguishes XLA compiler\u002FHLO launch divergence and host data-input stalls from TPU chip, SparseCore, ICI, or network fabric faults. Use when multi-slice TPU training jobs freeze without progressing steps, emit `HANG_DETECTED`, or stall in collective operations. Don't use for gradual step-time throughput drops without hangs (use gke-ai-troubleshooting-tpu-performance-degradation) or pod preemption\u002Feviction restarts (use gke-ai-troubleshooting-jobset-interruption).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-mxla-hang\u002FSKILL.md",[],[443,347,209,349,416,324,413,444,417,418,352,353,445,446,307,359,421,447,448,249,449,450],"gke ai troubleshooting tpu mxla hang","diagnostics","network","data","input","collective","Preemption","Training Jobs",{"skillNo":452,"name":453,"nameZh":18,"description":454,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":455,"categories":456,"keywords":457,"publishedAt":228},"01M6992ZX02SEBJQVRJH4KZE87","gke-ai-troubleshooting-tpu-metrics-monitoring","Monitors and troubleshoots GKE TPU workloads, nodes, and node pools using GKE system metrics and PromQL. Use when monitoring TensorCore duty cycle, TPU memory, node readiness, multi-host TPU node pool availability, host maintenance or preemption interruptions, and calculating MTTR or MTBI metrics for GKE TPUs. Don't use for general non-TPU GKE workload monitoring or non-metric TPU debugging.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-metrics-monitoring\u002FSKILL.md",[],[458,347,375,353,209,349,459,326,460,461,462,216,306,419,359,463,464,226,337,249],"gke ai troubleshooting tpu metrics monitoring","metrics monitoring","debugging","mttr","system","multi","availability",{"skillNo":466,"name":467,"nameZh":18,"description":468,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":469,"categories":470,"keywords":471,"publishedAt":228},"01M6992Y037TMWAB3Z885V2J83","gke-ai-troubleshooting-tpu-dynamic-slices-monitoring","Monitors, troubleshoots, and manages GKE TPU Dynamic Slices custom resources. Use when checking TPU slice lifecycle states, troubleshooting slice provisioning failures, validating single-slice or multi-slice (JobSet) workload manifests, or safely patching stuck finalizers and disabling the slice controller. Don't use for generic GKE cluster node pool creation or standard non-TPU workload management (use gke-basics or gke-cluster-creation instead).","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-tpu-dynamic-slices-monitoring\u002FSKILL.md",[],[472,347,353,209,349,212,262,473,474,355,259,222,288,210,434,308,226,391,475,249],"gke ai troubleshooting tpu dynamic slices monitoring","controller","patching","generic",{"skillNo":477,"name":478,"nameZh":18,"description":479,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":480,"categories":481,"keywords":482,"publishedAt":228},"01M6992WJD7EMBMR6AAFZS8FTM","gke-ai-troubleshooting-node-unresponsive-timeout","Diagnose and mitigate GKE TPU or GPU nodes stuck in NotReady \u002F NodeStatusUnknown (\"Kubelet stopped posting node status\") due to host kernel panics, hardware lockups, or disabled node auto-repair. Use when nodes stop heartbeating beyond the node auto-repair threshold and pods remain stuck in Terminating. Don't use for healthy nodes, pod-only application crashes, or routine GKE upgrades.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-node-unresponsive-timeout\u002FSKILL.md",[],[483,209,349,355,484,212,485,486,347,487,392,217,488,489,350,490,359,226,491,249],"gke ai troubleshooting node unresponsive timeout","timeout","kernel","gke upgrades","status","only","stop","posting","routine",{"skillNo":493,"name":494,"nameZh":18,"description":495,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":496,"categories":497,"keywords":498,"publishedAt":228},"01M6992TQ9V7XW5BB9F3Q9NNW1","gke-ai-troubleshooting-jobset-interruption","Diagnoses GKE JobSet interruptions, restarts, and preemptions for AI\u002FML training workloads autonomously. Use when troubleshooting JobSet restart loops, spot VM preemptions, node readiness failures, host VM issues, or coordinator worker crashes. Don't use for general GKE cluster creation, basic workload deployment, or non-JobSet application issues.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-jobset-interruption\u002FSKILL.md",[],[499,209,349,212,262,417,263,215,355,306,500,501,222,502,359,210,503,504,226,249],"gke ai troubleshooting jobset interruption","restart","worker","basic","loops","coordinator",{"skillNo":506,"name":507,"nameZh":18,"description":508,"descriptionZh":18,"repositoryFullName":201,"ownerLogin":164,"ownerAvatarUrl":166,"ownerAccountNo":163,"ownerDisplayName":165,"repositoryNo":202,"starsCount":203,"licenseSpdxId":204,"licenseFileUrl":18,"sourceHtmlUrl":509,"categories":510,"keywords":511,"publishedAt":228},"01M6992S3G6YRFRG97SABMXHSP","gke-ai-troubleshooting-handle-disruption-gpu-tpu","Diagnoses, predicts, and mitigates node disruptions during Compute Engine host maintenance and hardware or software maintenance events for GPU and TPU workloads on GKE. Use when diagnosing node disruptions, predicting host maintenance events on GPU\u002FTPU nodepools, inspecting node interruption PromQL metrics, auditing node taints, or configuring workload protection strategies (graceful termination, opportunistic maintenance, PodDisruptionBudgets). Don't use for general GKE cluster creation, network policy configuration, or non-disruption workload deployment.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-ai-troubleshooting-handle-disruption-gpu-tpu\u002FSKILL.md",[],[512,347,513,209,349,514,515,263,393,216,516,306,220,517,222,359,210,226,337,249],"gke ai troubleshooting handle disruption gpu tpu","gpu","policy","gke cluster creation","maintenance","strategies",248]