Dynamo Router 启动器：拉起 Dynamo 路由并做端点冒烟验证
Dynamo Router 启动器（@org-02qudk26/dynamo-router-starter，v1.0.0）是 SkillHub it-ops-security 类企业 Skill，用于启动或修补 NVIDIA Dynamo 的 round-robin、KV、least-loaded、device-aware 等路由模式，并通过 check_router_health.py 校验 /v1/models 与 chat 端点。文章说明前置条件、skillhub 安装到 Cursor/Claude/Codex skills 目录的方法，以及本地 CLI 与 Kubernetes env 修补的典型命令，并区分与 dynamo-recipe-runner、dynamo-troubleshoot、dynamo-benchmark 的分工。
Read Moregke-golden-path: Default Configuration and Standard Path for GKE Production Clusters
gke-golden-path is an Agent Skill in Google's official google/skills repository (v1.1.0, Containers) that provides Autopilot golden path defaults, Day-0/Day-1 rules, and guardrails for designing GKE clusters, production readiness checks, and configuration audits. It covers Always-Apply items such as private nodes, Workload Identity, monitoring and logging, and PSS restricted, and outlines four recommendations for upgrade disruption: PDBs, replicas, probes, and Surge. Install it with npx skills add google/skills; for workload scaling, use gke-workload-scaling.
Read MoreGKE Basics: Using Agent Skill to Avoid Common Pitfalls in GKE Cluster Creation and Workload Deployment
gke-basics is a Containers-category Skill (v1.1.1) in the google/skills repository, designed to guide agents through creating GKE clusters, obtaining kubectl credentials, choosing between Autopilot and Standard modes, and deploying workloads. It also documents common pitfalls, including private clusters and Autopilot CPU request increments. Topics such as Workload Identity, networking, security, and upgrades are routed to dedicated Skills in the same series. It can be installed with `npx skills add google/skills` and used alongside the CLI, MCP, and IaC documentation under references.
Read Moredsh-wsl-helm: A Read-Only Helm Tool for DeepSeek Harness
The `dsh-wsl-helm` plugin provides read-only access to Helm Releases in Kubernetes environments for DeepSeek Harness (DSH). It was developed by user 173787247 and focuses on viewing release lists, status, and history, while prohibiting install or uninstall operations. The plugin provides four core tools: `helm_list`, `helm_release_status`, `helm_history`, and `helm_status_tool`, and supports permission management by configuring `allowedContexts`. The installation command is `dsh plugin --profile web add github:17378`
Read MoreDSH AIOps: An Intelligent Operations Plugin for Kubernetes
`dsh-aiops` is a DSH workflow plugin that integrates alert ingestion, fault diagnosis, and observability capabilities, aiming to address the fragmentation of operations toolchains. It is Alertmanager-driven and provides core features such as authentication, session routing, skill package diagnosis, and structured reporting. Installation requires meeting the version requirements for DSH, Node.js, and pnpm, and configuring `allowBuilds` in `pnpm-workspace.yaml` correctly. Configuration includes environment variables (Prometheus, Alertmanager, Kubeconfig, etc.) and Webhook listening. The plugin provides an AIOps Portal in the DSH Web interface and only reads `AIOPS_WORKSPACE`.
Read Morejackson-chen97/dsh-devops: GitLab and Kubernetes DevOps Console
The DeepSeek Harness DevOps (DSH) plugin is designed to address efficiency issues in interactions between AI agents and CI/CD systems as well as Kubernetes clusters. It provides a visual Web console that supports GitLab project management (MRs and pipelines) and Kubernetes cluster operations (Pod status, logs, and rollbacks), with built-in background monitoring and alerting. It also supports multi-server configuration and internationalization. Installation via local checkout is recommended. Once configured, the plugin can be used interactively through conversation to directly invoke AI tools for automated operations.
Read Moredsh-kubectl: kubectl Integration Plugin for DeepSeek Harness
dsh-kubectl is an open-source plugin that encapsulates kubectl capabilities into the DeepSeek Harness (DSH) framework, providing agents with structured, confirmation-based interactions for Kubernetes operations. The plugin supports querying resources, viewing logs, port forwarding, and executing write operations with user confirmation, addressing the issue that conventional output is often not agent-friendly and can be parsed unreliably. Installation requires using the patch command and configuring the kubectl path, while write operations rely on explicit confirmation to prevent accidental execution.
Read MoreUsing kubernetes-deploying Skill: Let AI Agents Generate Kubernetes Deployment Manifests According to Specifications
### 翻译结果： `kubernetes-deploying` from spencerpauly/awesome-cursor-skills is an Agent Skill designed for Kubernetes deployments. It encapsulates best practices including Deployment, Service, Ingress, ConfigMap, Secret, health checks, release strategies and HPA into the SKILL.md file, enabling AI programming tools such as Cursor, Claude Code and Codex to generate standardized manifests and work with common kubectl commands. This article introduces its capabilities, installation and activation methods, typical usage scenarios and usage notes based on official documentation.
Read MoreGateway API v1.6: TCPRoute and UDPRoute Promoted to Standard, L4 Routing Generally Available
Kubernetes Gateway API v1.6.0 (released on June 30, 2026) has promoted TCPRoute and UDPRoute to the Standard channel, with L4 TCP/UDP routing reaching GA status. The API has been migrated to gateway.networking.k8s.io/v1, and v1alpha2 has been deprecated. This article sorts out L4 routing configuration examples, the unified evolution path with Ingress/Service Mesh, as well as the experimental API group gateway.networking.x-k8s.io and XBackend egress capabilities, for reference of cluster network upgrades.
Read MoreKubernetes 1.37 Coming Soon: HPA Scale-to-Zero, DRA Graduation to Stable, IPVS Mode Phased Out
Kubernetes 1.37 is scheduled for release on August 26, 2026, and the official Sneak Peek and enhancement tracker list 86 changes. This article focuses on three areas that require the most attention from operations teams: HPAScaleToZero has entered Beta and is enabled by default, allowing queue-based workloads to natively scale to zero; DRA device taint tolerance (KEP-5055) has reached GA, enabling GPU clusters to taint and evict pods based on specific devices; the kube-proxy IPVS mode has started the deprecation process, it will be disabled by default in 1.40 and completely removed in 1.43, so teams should migrate to nftables in advance. It also sorts out pre-upgrade check items such as metrics.k8s.io GA, containerd 2.0 and cgroup v2 upgrades.
Read MoreK8s-native AI Inference: The Ecosystem of Kthena, Grove and KAI Scheduler
In 2026, the Cloud Native Computing Foundation (CNCF) Community Accelerate will elevate LLM inference to a first-class citizen in Kubernetes. The CNCF Volcano subproject Kthena offers KV Cache-aware routing and Prefill-Decode separated scheduling; NVIDIA open-sourced Grove and KAI Scheduler to support topology-aware gang scheduling for multi-node inference; the CNCF Sandbox project KAITO simplifies vLLM model deployment. Based on cross-verification of official blogs and GitHub repositories, this article sorts out the responsibility boundaries and collaboration modes of the three projects across model deployment, traffic routing, and GPU scheduling layers, and provides a minimal YAML example for PD separation, providing a reference for platform engineering teams to build a unified K8s AI infrastructure layer.
Read MoreKubernetes 1.37 Is Coming: A Comprehensive Guide to DRA Enhancements, AI/ML Batch Scheduling and nftables Transition
Kubernetes v1.37.0 is scheduled for release on August 26, 2026, with approximately 86 enhancements in its milestone. Based on the official release calendar and SIG Release Highlights, this article discusses the significance of DRA device taints/extended resources GA, Workload API Beta and CompositePodGroup Alpha for AI/ML batch scheduling, interprets the gradual transition of kube-proxy to nftables (KEP-5343 warning mechanism), Pod certificates GA and Kubelet Rootless Beta, and provides upgrade suggestions such as kube-proxy explicit configuration, Feature gate and API deprecation.
Read More