Looking for a way to simplify deploying LLMs on Kubernetes? This project provides everything you might need.

llmaz is an inference platform that integrates various Open Source projects for running LLMs. It supports:

- Different inference backends: vLLM, llama.cpp, Ollama, Text-Generation-Inference, SGLang, and TensorRT-LLM.
- Different model providers: HuggingFace, ModelScope, and ObjectStores.
- Chatbot interface based on Open WebUI.
- Heterogeneous devices.
- Distributed inference via multi-host and homogeneous xPyD support with LeaderWorkerSet.
- Envoy AI Gateway for token-based rate limiting, model routing, and more.
- Horizontal Pod scaling (HPA) and node autoscaling (Karpenter).

▶️ GitHub repo

Language: Go | License: Apache 2.0 | 278 ⭐️

#tools #aiml

If you’re not overwhelmed yet with your work after a holiday break… or if you’re just into having some educational and practical fun with Kubernetes, don’t miss this project!

K8sQuest is a new gamified training platform for K8s, where you need to troubleshoot and fix various issues using a GUI terminal featuring arcade game styling. Importantly, it can be self-hosted locally. The project comes with:

- 50 challenges, 5 categories, 3 difficulty levels;
- different K8s topics covered, including basics, scaling, networking, storage, and security;
- progressive hints and step-by-step guides, points for completed challenges, and progress auto-saving.

▶️ GitHub repo
💬 Reddit announcement

Language: Shell, Python | License: Apache 2.0 | 326 ⭐️

#tools #fun #career

Feel a need to validate your Dockerfiles against best practices? Consider trying this new tool.

Dockadvisor is a fast linter for Dockerfiles that helps you keep them optimal and consistent. Here’s what it offers:

- applying 60+ rules that cover standard instructions (FROM, RUN, ENV, etc.) and multi-stage builds;
- performs security checks, such as specifying secrets in variables;
- scores the quality of your Dockerfile on the scale from 0 to 100;
- can be used as a web interface, Go library, or WebAssembly module (i.e. executed in the browser).

▶️ GitHub repo

Language: Go | License: Apache 2.0 | 70 ⭐️

#tools

As many are aware, MinIO has recently gone to "maintenance mode.” If you’ve been looking for a Kubernetes-friendly alternative, check out this new project.

Garage, an S3-compatible distributed object storage, just got an unofficial Kubernetes operator. Still being in its alpha, it simplifies deploying and maintaining Garage clusters with the following features implemented:

- Deploying StatefulSets with proper configuration, storage, and networking;
- Bucket creation with quotas;
- S3 key management with automatic credential generation;
- Multi-cluster federation by connecting Garage clusters across K8s instances.

▶️ GitHub repo
💬 Reddit announcement

Language: Go | License: Apache 2.0 | 44 ⭐️

#tools #storage #news

Kubernetes-based alternatives to Heroku are real. Here’s one of them.

Canine positions itself as a “developer-friendly PaaS for your Kubernetes”. It’s focused on small development teams and simplifies using Kubernetes for them by providing:

- container builds performed via Docker BuildKit or Buildpacks;
- automatic deployment to GitHub and GitLab;
- web UI to deploy, scale, and manage (e.g., configure resource constraints) apps running in Kubernetes;
- integration with existing K8s tools, such as Helm, cert-manager, and Telepresence;
- single sign-on via SAML, OIDC, and LDAP.

▶️ GitHub repo

Language: Ruby | License: Apache 2.0 | 2716 ⭐️

#tools #gui

vCluster introduced vind, marketed as a better kind

vCluster Labs (previously known as Loft Labs) released a new tool called vind (vCluster in Docker). It is built on top of vCluster and allows you to run Kubernetes clusters directly as Docker containers, similarly to what kind (Kubernetes IN Docker) offers. However, it comes with the following extra features:

- pausing the clusters when they're not in use and resuming them;
- automatic LoadBalancer support;
- image caching (pull-through cache via local Docker daemon);
- support for connecting external nodes, which can be real cloud instances;
- support for choosing CNI and CSI plugins;
- built-in vCluster Platform UI.

You can find more details about vind on GitHub and in yesterday’s video presentation on LinkedIn.

#news #tools

Optimising resources in Kubernetes is something we all want to do at some point. This new project aims to assist in that.

CruiseKube, dubbed as “Autopilot for Kubernetes”, is a controller that watches your K8s workloads and adjusts the resources accordingly. Here’s what it does:

- Continuously evaluates current CPU/memory usage and updates resource requests.
- Considers CPU pressure (PSI metrics) and other Pods on the node when resizing.
- Watches OOM memory values in stats and triggers Pod eviction when needed.
- Uses Prometheus as the primary metrics source.
- Provides a web UI to see and manage your settings.

▶️ GitHub repo
💬 Reddit announcement

Language: Go | License: MIT | 48 ⭐️

#tools

Node Readiness Controller for Kubernetes

Last week, a new Kubernetes SIG project was announced. The Node Readiness Controller can be used to define additional requirements for node readiness (e.g., GPU drivers are loaded). The controller will manage node taints to prevent scheduling until the required conditions are satisfied. It supports bootstrap-only and continuous enforcement modes. Currently, the project is in its alpha.

Find more details in the project’s documentation and on GitHub.

#news #tools

This new UI aims to ensure “modern Kubernetes visibility” by providing comprehensive information on your cluster and its workloads, along with several management features.

Radar is a dashboard that is intended to be “blazing fast”, displays real-time information, and runs as a single binary, not requiring to be installed on a cluster. It comes with:

- General cluster overview, including the stats for existing resources, resource utilisation, and unhealthy workloads.
- Detailed interactive graphs for Kubernetes resources with their full hierarchy and an image filesystem viewer for Pods.
- Live network traffic visualisation (via Hubble or Caretta).
- Timeline of Kubernetes events and resource changes.
- Management for Helm releases and GitOps (Argo CD and Flux) resources.
- Automatic discovery of CRDs and integrations for Gateway API, Karpenter, KEDA, cert-manager, Prometheus Operator, and Trivy.
- MCP server for AI integration.

▶️ GitHub repo

Language: TypeScript, Go | License: Apache 2.0 | 863 ⭐️

#tools #gui

Watching your Kubernetes Pods in real-time 3D space sounds like a deal for Friday, doesn’t it? 🙃

Observatory is a visualisation dashboard that makes this possible. Originally built for K3s, it works with other Kubernetes distros as well, allowing you to watch your containers like never before. What it offers:

- Displaying your Kubernetes nodes and Pods in the 3D space where you can travel;
- Showing sidecars as orbiting moons for multi-container Pods;
- Providing the current and continuously updated state of Pods (running, pending, etc.) as well as their memory and CPU usage visualised as size and colours.

▶️ GitHub repo
💬 Reddit announcement

Language: Go, TypeScript | License: GPL v3 | 28 ⭐️

#tools #gui

Interested in seeing the contents of your container images without running them? Check out this new tool.

cek (container exploration kit) is a CLI tool for exploring the OCI images filesystem. Unlike Skopeo, it works with the container itself (rather than the container registry), i.e. it can read images directly from Docker, Podman, or containerd in addition to pulling them from remote registries. cek allows you to:

- list files in your image and display the directory tree structure;
- read file contents;
- inspect image metadata;
- export images to tar files.

▶️ GitHub repo
💬 Reddit announcement

Language: Go | License: MIT | 261 ⭐️

#tools #storage

Homelabs are a fun way for engineers to learn, experiment, and innovate. Sharing such setups can bring even more benefits to a wider community! Here’s one of such repos you can try this weekend ;)

Homernetes is a Kubernetes cluster for a homelab based on Talos and Proxmox. It features an automated 8-step bootstrap to provision a cluster on bare metal using Terraform. What else does it have?

- GitOps-driven approach based on Argo CD;
- Preloaded randomly-generated passwords/secrets for all services with Vault;
- Networking with encryption and observability based on Cilium;
- Metrics and logs based on Prometheus, Grafana, and Loki;
- cert-manager to handle certificates, Harbor as container registry, CNPG with PostgreSQL used for internal services, and more.

▶️ GitHub repo
💬 Reddit announcement

License: GPL 3.0 | 142 ⭐️

#tools #IaC #gitops

Ever noticed that unused resources, such as Secrets and ServiceAccounts, might pile up in your Kubernetes cluster? There is a tool that removes them.

K8s cleaner is a controller that finds stale and unhealthy resources and removes or updates them. Here’s what it offers:

- Identifying various types of unused Kubernetes resources, including ready-to-use examples for ConfigMaps, Secrets, Roles/ClusterRoles, ServiceAccounts, PVs/PVCs, Deployments, and StatefulSets;
- Identifying resources based on annotations for maximum lifespan or expiration date;
- Using Lua scripts to define custom selection criteria;
- Scheduling the scans for finding and removing/updating unused resources;
- Notifications via emails, Slack, Discord, Teams, Telegram, etc.;
- Web UI showing existing issues, cleaners, and Lua scripts.

▶️ GitHub repo

Language: Go | License: Apache 2.0 | 755 ⭐️

#tools

This project is a Kubernetes operator that implements a so-called health-check platform.

Canary checker is a monitoring tool that includes 35+ ready-to-use health checks for various Kubernetes workloads. It works as a K8s operator that has no dependencies and offers:

- checks for popular protocols (HTTP, DNS, ICMP, TCP), databases (SQL-based, MongoDB, Redis, Elasticsearch), infrastructure components (such as Kubernetes resources and EC2 instances), file systems (NFS, S3, SMB), and more;
- alert aggregation from Prometheus, AWS CloudWatch, and Dynatrace;
- integration testing support for JMeter, JUnit, K6, Newman and Playwright;
- Prometheus exporter replacement with custom metrics from the check’s results;
- built-in Web dashboard displaying existing checks and their stats.

▶️ GitHub repo

Language: Go | License: Apache 2.0 | 326 ⭐️

#tools #observability

Can luring random online attackers into your Kubernetes clusters be fun or even useful? At least, we have a tool to try it out with minimal effort.

Koney is a Kubernetes operator that implements so-called “deception policies” to discover and deter malicious users. Currently, it supports honeytokens as the only deception method by:

- creating fake “sensitive” files, such as /run/secrets/koney/service_token, in all selected Pods;
- monitoring the attempts to access them by collecting alerts from the Cilium’s Tetragon operator or via a smaller eBPF-powered file access monitoring tool (kivebpf);
- logging these events and, optionally, sending alerts to external systems (currently, it supports only Dynatrace).

Future project plans include adding other deceptive techniques (HTTP endpoints and payloads) and integrating the operator with Kyverno policies.

▶️ GitHub repo

Language: Go | License: AGPL 3.0 | 89 ⭐️

#tools #security

Upbound introduced Modelplane for AI inference

The company behind Crossplane, a CNCF project that graduated at the end of 2025, announced its new Open Source project called Modelplane. The CNCF's CTO already mentioned on LinkedIn that this project would be a great fit for CNCF.

Basically, Modelplane is a control plane for AI inference built on Crossplane, which allows you to run any model on any serving engine on any infrastructure. Here's its brief description from the official announcement:

Modelplane sits above your inference clusters and operates them as one inference platform, reconciling the whole fleet toward the state you declare. It does for the fleet what Kubernetes does for the cluster: platform teams and developers describe what they want as Kubernetes resources, and Modelplane composes the clusters, places the models, and exposes the endpoints to match.

Additional resources include the getting started documentation and the GitHub repo.

#news #tools #aiml

William Rizzo, a CNCF Ambassador and Global Field CTO at Mirantis, announced his project, KubeSwift, which provides Kubernetes-native VM orchestration on Cloud Hypervisor.

Here’s the idea behind KubeSwift: “KubeSwift runs virtual machines as Kubernetes workloads. You define a VM with a custom resource; controllers reconcile it into a Pod; inside that Pod, swiftletd launches a hypervisor.” (Cloud Hypervisor, on which this project relies, is a VMM (Virtual Machine Monitor) for cloud workloads written in Rust.)

While still being in the pre-1.0 state, KubeSwift already supports disk boot from cloud images and direct kernel boot from OCI artifacts, Multus and OVN-Kubernetes for networking, exposing guest ports as Kubernetes Services, GPU passthrough, disk snapshots, offline and live migrations, managing fleets of identical VMs, swiftctl CLI tool for lifecycle management and operations, and many other features.

▶️ GitHub repo

Language: Go | License: AGPL 3.0 | 18 ⭐️

#news #tools

Thought of running agents on Kubernetes more efficiently? Here’s a project from Google that you might find useful.

Agent Substrate is an experimental system for running agents and similar workloads on Kubernetes at scale, providing specialised scheduling and controls to improve scalability and reduce latency. It relies on Kubernetes Pods and Pod autoscaling (but not K8s control plane) and manages the lifecycle of "actors" (agents or similar apps): creating and destroying them, assigning them to Pods, and routing traffic to them. It handles OCI containers at the kernel level via gVisor, so it can work with apps that are built on any stack.

▶️ GitHub repo

Language: Go | License: Apache 2.0 | 912 ⭐️

#tools

Another kube-oidc-proxy fork is alive, featuring multi-issuer authentication to managed Kubernetes clusters.

This new project, created by Rafal Pieniazek, implements a reverse proxy that enables authentication from multiple OIDC issuers at once to your managed K8s clusters, such as EKS, GKE, and AKS. To achieve this, the proxy sits in front of the API server, validates the bearer token against required OIDC issuers, and maps the token's claims to a Kubernetes identity. Then, impersonated headers for the mapped user are added, so that existing RBAC stays authoritative.

P.S. This project originates from kube-oidc-proxy by Tremolo Security, which itself is an active fork of the eponymous project by Jetstack.

▶️ GitHub repo

Language: Go | License: Apache 2.0 | 17 ⭐️

#tools #security

We all know (and many of us use daily) k9s as a truly flagship CLI tool for Kubernetes. But what if someone were to reimagine it today?

Meet sofka, which reimplements a well-known tool using one generic object pipeline (instead of separate structs for each resource type) and Rust. It performs better and offers lots of features you might need, including:

- Powerful keyboard-based navigation with curated columns, custom views, drill-down navigation, horizontal scrolling, global fuzzy find, compact mode, and mouse support.
- Powerful metrics and health support, covering Pod health, live CPU and memory values, workload health at a glance, evidence-based explanation for unhealthy resources, hierarchical Xray view, watch notifications, and more.
- Extended diagnostic capabilities with incident bundles and snapshots capturing the current table view.
- Embedded Helm and GitOps integration.
- Various executable actions, such as file transfers, debugging containers and Pods, exploring PVC contents, and right-sizing requests.

What makes it different from k9s? Built-in Flux CD and Argo CD support, port-forwarding running in the background, bulk actions with multiselect, skins (12 are available), and more.

▶️ GitHub repo
 
Language: Rust | License: Apache 2.0 and MIT | 672 ⭐️

#tools #CLI