Terraform modules for a GPU/LLM inference platform on Amazon EKS: Karpenter GPU NodePools, NVIDIA device plugin + DCGM, S3 model store via Mountpoint, vLLM with KEDA queue-depth autoscaling

Self-hosted Google Calendar mirror: busy blocks, rule-filed copies, meeting checks, archive and cleanup

Proxmox LXC + Terraform automation for a highly available K3s cluster with kube-vip

The browser's way in to Kubernetes: login, API access and live krm-stream resources.

APEX-OS Business Platform — Unified business operations platform: 30 modules (CRM, accounting, inventory, BI, IoT, compliance...), 500+ FastAPI endpoints, 58-page React dashboard

Prometheus exporter for the Komodor public API: reliability risks, issues and clusters as metrics, with a Grafana dashboard, Helm chart and Docker Compose demo.

Secure OpenAI-compatible LLM gateway with JWT auth, rate limiting, caching and prompt guard. FastAPI, Redis, Kubernetes.

Generic Helm chart for any Kubernetes workload: Deployment, Service, Ingress or Gateway API HTTPRoute, HPA. One chart for web apps, APIs, workers and stateful services.

☁️ Automated self-service application deployment platform inspired by Render & Railway for containerizing and deploying GitHub repos to Kubernetes.

Top Gitops Continuous Delivery (Opensource) Star if you like it!

Vendor-neutral health and readiness checks for agents, MCP, A2A and their dependencies. CLI, Go engine and experimental AHP serving.

CODEBERG MIRROR — Plethos is a self-hosted platform that lets organizations provide their members with isolated Kubernetes clusters on demand. Built on vCluster by vCluster Labs, it enables users to create and manage their own virtual clusters.

A hands-on Kubernetes course: tasks, hints, and broken manifests to debug. You write the YAML.

Production automation scripts for Linux, Cassandra, Kubernetes, database operations, backup/restore, migration, and operational tooling.

Regional Harbor stuck mid-removal

Kubernetes operator in Go for the WebApi resource: validation, rollouts and status.

crony：AI Agent 化运维平台（自愈 / 放量安全栈 / 评测 / 写动作审批与审计；含 K8s 执行后端与集群只读诊断）

A k3s cluster for Oracle Cloud Always Free Tier (AMD VMs) 🚀

Independent experimental Nebius GPU infrastructure cookbook and Backstage estate portal. Synthetic demo, Terraform recipes and dated sandbox evidence.

Top Virtual Kubernetes Clusters (Opensource) Star if you like it!

Cloud-native Go API deployed on a multi-node Kubernetes cluster using Kind, with Prometheus for metrics collection and Grafana for visualization.

Kubernetes incident-to-fix agent: triage → root cause → patch → sandbox verify → PR with evidence. Never merges — a human reviews.

Go library, native CLI, and offline storage decision workbench for Kubernetes, observability, and AI data workflows.

High-throughput distributed LLM serving cluster orchestrating vLLM and NVIDIA Triton with continuous batching and PagedAttention.

A lab exercise to see Helm in action on a local Minikube cluster. It draws from the `spring-microservices-docker-kubernetes` (applications) and `spring-microservices-gitops` (Kustomize manifests + Argo CD) ecosystem, but has been rewritten using Helm and features smaller-scale applications.

Relay verification codes and sign-in links from IMAP mailboxes to Telegram, with a web admin UI · 即時把信箱裡的驗證碼與登入連結轉發到 Telegram

Host security agent for Raspberry Pi and ARM Linux. Ten attack categories, ranked, as Prometheus metrics, JSON events and push alerts. One static binary, standard library only.

Read-only reverse proxy that serves static sites from S3-compatible buckets in Kubernetes: GET and HEAD only, host-based routing, YAML config, optional TLS with HTTP/2 and HTTP/3. Go, distroless.

The ork CLI and the OrkStage scan engine. Turns Terraform, CloudFormation, Kubernetes, Helm and Compose projects into OrkStage templates, then pushes and deploys them. Rust, also compiled to WebAssembly. Preview.

Dashboard of vulnerabilities, outdated container images and Helm charts (with subcharts) across Kubernetes clusters.