minikube headlamp plugin 0.3.0

A UI for managing local minikube clusters. This is an early release (0.3.0). Feedback welcome.

Thanks @gambtho and @illume

Added

• Bundled latest minikube version: v1.38.1
• HyperV driver support for Windows including elevated privileges, auto-enable, and stop/delete wrappers
• vfkit driver support for macOS
• ClusterStatus component on the home page to start and stop clusters
• Loading indicator while detecting available drivers
• Free disk and free memory warnings when creating a cluster
• Container runtime and CNI flags for cluster creation
• Detection of existing minikube profiles

Changed

• Renamed cluster creation page title to "Create Local Cluster"
• Lowered default RAM for new clusters to 3072 MB to support lower-memory machines
• Error banner shown when a command fails; output capped at 200 lines to prevent slowdowns
• Cancelling a running command now terminates the background process and no longer reloads the page
• Cluster name is now validated before being passed to the minikube CLI
• Fallback system info detection on Windows when wmic is unavailable

Fixed

• Fix driver select not rendering and value prop deprecation
• Fix cluster name input so the initial value can be edited
• Fix form label overlapping the cluster name input field
• Fix Windows path detection for the plugin binary
• Fix HyperV stop and delete commands
• Fix minikube profile list detection

Kubeflow - 0.1.0-alpha

First public release. This is an alpha release. APIs and interfaces may change before v1.0. Feedback and bug reports are welcome.

About

Kubeflow is an open-source machine learning platform for Kubernetes. It encompasses Notebook Servers, ML Pipelines, AutoML (Katib), distributed Training Operators, and Spark Operators, all managed through Custom Resource Definitions.

Headlamp is an open-source, extensible Kubernetes web UI designed to be extended via plugins. Headlamp plugins are first-class citizens: they hook into the sidebar, routing, and resource rendering system to deliver custom resource views that feel native to the UI.

Managing these resources often requires the Kubeflow Central Dashboard or raw kubectl commands. This plugin brings deep, Kubernetes-centric observability directly into Headlamp, complementing the official dashboard with SRE-level insights.

20260409-1207-47.1448485.mp4

Bringing Full Kubeflow Notebook Observability and Troubleshooting Directly Into Headlamp

This release empowers platform operators with specific workflows that were previously difficult to perform without the CLI:

1. Troubleshooting Notebook Failures: When a notebook stays "Pending" in the official dashboard, operators can now use the Notebook Detail Conditions to immediately identify exact failure reasons like ImagePullBackOff, Insufficient GPU capacity, or Taint/Toleration mismatches.
2. Cluster-Wide Resource Auditing: Use the Resource Aggregation Cards and Images in Use table to identify over-provisioned namespaces or find expensive GPU-hoarding notebooks across the entire cluster.
3. Security & Injection Validation: Verify that sensitive credentials (AWS/GCP/Database secrets) are correctly propagating by using the PodDefault Injected Environment Variables view, confirming admission controller rules without manually inspecting pod YAML.

---

Features in v0.1.0-alpha

Notebooks Dashboard Overview

The landing page for Notebooks, featuring resource aggregation cards (CPU/Memory/GPU), image distribution tables, and namespace-level statistics for all running servers.

---

Notebook Servers List and Deep Detail

Rich list and detail views for Notebook resources. Features include auto-detection of notebook types (Jupyter/VS Code/RStudio), detailed resource request tracking (including GPU), volume mount inspection, and direct access to pod logs.

The detail view provides a deep dive into container specs, status conditions with real-time reasons, sidecars, and tolerations.

---

Multi-Tenant Profile Management

Full visibility into Profile ownership, resource quotas (CPU/Memory/GPU), and associated plugins.

---

Integrated Sidebar Navigation

The plugin registers core Kubeflow resources under a dedicated section in the Headlamp sidebar. Routes are wrapped with automated CRD discovery to ensure a graceful UI even if only a subset of Kubeflow components are installed.

---

PodDefault Admission Inspection

Deep inspection of admission controller injection rules. Shows exactly which environment variables, volumes, and annotations are being applied to notebook pods, including source resolution for Secrets and ConfigMaps.

---

Compatibility

Component	Version
Headlamp	>= 0.13.1
Kubeflow Notebooks API	kubeflow.org/v1
Kubeflow PodDefaults API	kubeflow.org/v1alpha1
Kubeflow Profiles API	kubeflow.org/v1
Kubeflow Pipelines API	pipelines.kubeflow.org/v2beta1
Katib API	kubeflow.org/v1beta1
Training API	trainer.kubeflow.org/v1alpha1
Spark Operator API	sparkoperator.k8s.io/v1beta2

Known Limitations (Alpha)

• Notebook creation/deletion and start/stop are not yet supported via the UI.
• Pipelines, Katib, Training, and Spark have placeholder pages only; full views are planned for future releases.
• Provider-specific configurations (e.g., AWS S3 artifact stores) are not specialized in the UI yet.

Installation & Quick Start

For detailed setup instructions and local developer testing, please refer to the official README.

Quick start, try it out:

git clone https://github.com/headlamp-k8s/plugins
cd plugins/kubeflow
npm install
npm run build

Contributors

This plugin was developed by the Headlamp community.
@alokdangre

Feedback

This is an alpha release and community feedback shapes what comes next.

• Bug reports: Open an issue
• Feature requests: Start a discussion
• Contributing: CONTRIBUTING.md

Strimzi 0.4.0-alpha - Headlamp plugin for managing Kafka

Headlamp is a Kubernetes UI that runs in the browser or as a desktop app. Strimzi runs Apache Kafka on Kubernetes using custom resources (Kafka, KafkaTopic, KafkaUser, and related objects).

The Strimzi Headlamp plugin brings Strimzi workflows into Headlamp so operators and developers can inspect clusters, topics, and users, create and edit resources, watch status, and explore cluster topology—without switching to kubectl for day-to-day tasks.

Who it is for: platform teams running Kafka on Kubernetes, SREs debugging Strimzi resources, and anyone who already uses Headlamp and wants first-class Strimzi views next to the rest of the cluster.

Prerequisites: a Kubernetes cluster with the Strimzi operator installed, and Headlamp with this plugin loaded.

---

What you get in the UI

Area	Capabilities
Kafka clusters	List/detail, KRaft vs ZooKeeper mode, listeners & storage, status, topology modal (interactive graph)
Kafka topics	List, create, read, update, delete, filters (partitions, replicas, status, …)
Kafka users	SCRAM-SHA-512 and TLS auth, optional simple ACLs, view linked Secrets, delete with confirmation
Cross-cutting	Search and filters on lists, multi-namespace views, status aligned with Strimzi conditions

---

Demo — navigation and overview

headlamp-strimizi-demo.mp4

---

Screenshots — navigation and overview

Headlamp sidebar — Strimzi section

Plugin entry point: Strimzi with Kafka Clusters, Kafka Topics, Kafka Users.

Kafka Clusters — list view

Default list with namespaces, names, status, and mode (KRaft / ZooKeeper) where applicable.

Kafka Clusters — search / filter on list

Show search box and any active filters (status, mode, etc.).

---

Screenshots — Kafka cluster detail

Cluster topology — modal / graph (full view)

Interactive topology: brokers, controllers, ZooKeeper or KRaft layout, pools/pods as your cluster provides.

---

Screenshots — Kafka topics

Topics — list view

All topics with partitions, replicas, cluster label, status.

Topics — search / filter

Filter by status, partitions, replicas, or other supported fields.

Topics — create topic

Create dialog: name, partitions, replicas, retention, compression, etc.

Topic — edit

Edit flow updating partitions, retention, min ISR, or other fields.

Topic — delete confirmation

Confirmation dialog before delete.

---

Screenshots — Kafka users

Users — list view

List with authentication type, authorization, status.

Users — search / filter

Filter by auth type (SCRAM / TLS), status, etc.

Users — create user (SCRAM-SHA-512/TLS/ACLs)

Create user with different authentication.

User — view secret / password or cert

Flow showing View secret or equivalent for SCRAM password or TLS material.

User — delete confirmation

Confirmation before removing the user.

---

Install (short)

npm package: strimzi-headlamp

npm install strimzi-headlamp

Artifact Hub: strimzi-headlamp

Manual / tarball: extract under Headlamp’s plugins directory and restart Headlamp (see plugin README).

---

Try it with sample YAML

The repo includes test-files/ with example Kafka, KafkaTopic, and KafkaUser manifests and a short apply order—useful for demos and screenshots against a real cluster:

• strimzi/test-files/README.md

---

Call to Action

Please try the Strimzi Headlamp plugin in your own cluster and share feedback.

• Try it out: install the plugin, apply the sample manifests from test-files/, and manually test Kafka clusters, topics, users, and topology flows.
• Report bugs / request features: open an issue here: https://github.com/headlamp-k8s/plugins/issues
• Chat with us: join the Headlamp Slack channel and share your experience: Headlamp Slack
  (or replace with your project’s Slack invite/channel link)

We especially welcome:

• UX feedback (what felt confusing or slow)

Links

Resource	URL
Plugin README	https://github.com/headlamp-k8s/plugins/tree/main/strimzi
Issues	https://github.com/headlamp-k8s/plugins/issues
Strimzi docs	https://strimzi.io/documentation/
Headlamp	https://headlamp.dev/

Cluster API Plugin for Headlamp — v0.1.0-alpha

First public release. This is an alpha release. The interface will change before v1.0. Feedback and bug reports are welcome.

About

Cluster API is a Kubernetes sub-project that brings declarative, Kubernetes-style APIs to cluster lifecycle management. It lets platform teams provision, upgrade, and manage the lifecycle of Kubernetes clusters using standard Kubernetes objects — stored and reconciled in a management cluster. CAPI introduces custom resources like Cluster, Machine, MachineDeployment, MachineSet, and KubeadmControlPlane.

Headlamp is an open-source, extensible Kubernetes web UI designed to be extended via plugins. Headlamp plugins are first-class citizens: they hook into the sidebar, routing, and resource rendering system to deliver custom resource views that feel native to the UI.

Managing CAPI resources across a fleet has historically required raw kubectl commands or purpose-built CLIs(which is fine). This plugin brings full CAPI visibility directly into Headlamp.

---

Demo

Cluster-Apiv0.1.0.mp4

What This Plugin Does

The Cluster API Plugin for Headlamp (@headlamp-k8s/cluster-api) adds dedicated UI views for all core Cluster API resources to your Headlamp dashboard. You can browse clusters, inspect machines, track control plane health, scale deployments, and observe the full CAPI object graph — all without leaving the browser.

---

Highlights

• Full Cluster API visibility inside Headlamp (no kubectl required)
• Rich UI for clusters, machines, control planes, and deployments
• One-click scaling for MachineDeployment and MachineSet
• Dynamic API version support (v1beta1, v1beta2)
• Topology-aware UX with ClusterClass detection
• Clear parent–child relationships (Owned Machines)

---

Why This Plugin Matters

Managing Cluster API resources typically requires raw kubectl commands and deep familiarity with resource relationships.

This plugin brings:

• Visual clarity into cluster state
• Faster debugging of machines and control planes
• Simplified operations like scaling
• A unified UI for platform teams

—all directly inside Headlamp.

---

Contributors

This plugin was started from scratch by the Headlamp community. The initial plugin scaffold was created by a community contributor and later extended significantly for this release.

Contributor	Role
@mboersma	Initial plugin — c4f9267: cluster-api: Added plugin for Cluster API
@ChayanDass	Core development, architecture, UI/UX design, and feature implementation
headlamp-k8s org	Maintainers

A big thank you to the contributor who laid the foundation with the initial plugin commit. Without that starting point, this release would not exist.

---

Installation

Build from source:

git clone https://github.com/headlamp-k8s/plugins
cd cluster-api
npm install
npm run build

Requirements:

• Headlamp >= 0.13.1
• Node.js >= 18
• A management cluster with Cluster API installed (v1beta1 or v1beta2)

---

Compatibility

Component	Version
Headlamp	>= 0.13.1 (requires @kinvolk/headlamp-plugin ^0.13.1)
Cluster API	v1beta1, v1beta2
Kubernetes	1.26+
Node.js	18+

---

Features

Navigation & Sidebar

The plugin registers all Cluster API resources under a dedicated Cluster API section in the Headlamp sidebar. Resources are ordered logically for platform teams:

A CapiRouteWrapper component wraps all plugin routes and provides a consistent error boundary and loading state across all views. A plugin-level Dashboard component serves as the landing page.

---

Clusters

List view — displays all Cluster resources in the management cluster with control plane and worker replica status columns for enhanced visibility.

Detail view — shows full cluster spec, status conditions, infrastructure reference, control plane reference, and owned Machines.

---

Workload Resources

The plugin provides full support for workload-level resources:

• MachineDeployments — manage replica counts, rolling updates, and deployment strategies
• MachineSets — view replica status and ownership relationships
• Machines — inspect individual machine state, node references, and conditions
• MachinePools — manage pools of machines with provider-specific scaling

Each resource includes:

• List view with replica and status insights
• Detail view with full spec and conditions
• Ownership relationships and navigation

All resources follow a consistent UI pattern, ensuring a predictable and intuitive experience. Machines show node references, provider IDs, versions, and enhanced condition status rendering that clearly indicates warning/error states.

---

Control Plane Resources

KubeadmControlPlane — list and detail views show replica status, version information, rollout strategy, and associated control plane Machines. Dynamic API versioning ensures compatibility across v1beta1 and v1beta2 endpoints.

KubeadmControlPlaneTemplates — template resources with dynamic API versioning and loading states for control plane configuration.

---

Configuration Resources

KubeadmConfigs — list view includes a dedicated column for the Data Secret name (new in this release). Detail view shows full config spec, bootstrap data status, and owner references.

KubeadmConfigTemplates — template resources with dynamic API versioning, loading states, and template spec display.

---

Cluster Management

ClusterClasses — list and detail views show infrastructure template references, control plane template references, worker class definitions, and health check details.

MachineHealthChecks — view unhealthy conditions, max unhealthy thresholds, and node startup timeouts. Includes dynamic API versioning and improved loading states.

MachineDrainRules — manage machine drain behavior with API versioning support and loading states.

---

Owned Machines (Parent–Child Relationships)

Each resource clearly displays the machines it owns, making it easy to understand the Cluster API object hierarchy.

• MachineDeployments → MachineSets → Machines
• KubeadmControlPlane → Control plane Machines

This helps users:

• Quickly trace resource ownership
• Debug issues across the cluster hierarchy
• Navigate between related resources easily

The Owned Machines section is available in detail views and shows all child machines with relevant status and metadata.

---

Machine Templates & Infrastructure

Resource detail views expose:

• Bootstrap configs
• Infrastructure references
• Machine templates

This provides full visibility into how machines are provisioned and helps debug issues across the cluster lifecycle.

---

KubeadmConfig Spec & File Rendering

The plugin provides deep inspection of KubeadmConfig specifications, including inline file content and configuration details.

• Full KubeadmConfig spec rendered in a structured UI
• Inline file definitions (files) with:
  • Path, permissions, encoding
  • Rendered file content (YAML / config)
• Extra arguments and volumes clearly displayed
• Join and init configuration visibility

This allows users to:

• Inspect bootstrap configuration without accessing nodes
• Debug admission configs, kubelet args, and control plane settings
• Understand exactly what is being applied during node initialization

All configurations are presented in a readable, UI-friendly format, eliminating the need to manually inspect raw YAML or secrets.

Sca...

Volcano Plugin For Headlamp 0.1.0-alpha

Volcano is a cloud-native batch scheduler for Kubernetes, built for high-performance computing and AI/ML workloads.
This first release of the Volcano Headlamp plugin adds a visual interface for core Volcano CRDs so users can inspect scheduling state directly in Headlamp.
Included in this release:

• Volcano Jobs, Queues, and PodGroups list/detail views
• Volcano sidebar + route registration
• Cross-resource navigation (Job ↔ Queue/PodGroup, PodGroup ↔ Queue)
• Consistent status rendering and events integration
• Deployable test manifests for common job scenarios
  PR: #551

---

Demo

Volcano-Demo.mp4

---

Jobs (batch.volcano.sh/v1alpha1)

The Jobs view provides an operational overview of Volcano jobs and task execution.

• List view: status, queue, running/min-available, task count, age
• Detail view: summary info, Pod status section, Tasks section, Events
• Navigation: links to related Queue and PodGroup (when available)

---

Queues (scheduling.volcano.sh/v1beta1)

The Queues view surfaces queue-level scheduling configuration and allocation state.

• List view: state, weight, parent queue, age
• Detail view: capacity limits, allocated resources, events
• Navigation: parent queue link from queue details

---

PodGroups (scheduling.volcano.sh/v1beta1)

The PodGroups view highlights gang-scheduling state and progress.

• List view: phase, min member, running count, queue, age
• Detail view: progress, conditions, min resources, events
• Fallback behavior: clear message when conditions are not reported

---

Navigation and UX

This release introduces a dedicated Volcano section in the Headlamp sidebar and registers list/detail routes for all supported Volcano CRDs.
Status presentation is consistent across resources using shared status-to-color mapping.

---

Validation Assets

To make manual validation easier, this release also includes deployable manifests under volcano/test-files/deploy/:

• namespace.yaml
• queue.yaml
• job-running.yaml
• job-completed.yaml
• job-unschedulable.yaml
• job-failed.yaml

---

Feedback and Community

If you try the Volcano plugin, we’d love to hear how it works in your cluster!

• Report bugs or request features: https://github.com/headlamp-k8s/plugins/issues/new
• Join Kubernetes Slack: https://slack.k8s.io
• Chat with us in #headlamp on Kubernetes Slack

Knative Plugin For Headlamp 0.2.0-alpha

Knative provides powerful serverless capabilities to Kubernetes, such as scale to zero, revision-based deployment, traffic splitting, etc. While the kn CLI provides strong operational control, a visual interface is beneficial for many Kubernetes users.
The knative plugin for headlamp fills this very gap by providing an intuitive visual interface for managing Knative resources like Kservice, Revisions and Networking.

This allows users to Add/Edit Kservice, View Revisions, Manage traffic splitting between revisions, Autoscaling configuration and Shows Network Details

KService

A Knative Service (KService) is the top-level resource that manages the entire lifecycle of your serverless workload. It automatically provisions the underlying routes, configurations, and revisions needed to get your application running and exposed to traffic.

Since many fields may require frequent changes, we have added a toggle switch to switch between Read and Edit mode.

Traffic Splitting

Traffic splitting between all revisions are shown along with their tags which redirect to their respective URL

Autoscaling Configuration

Autoscaling features like Target Utilization %, Container Concurrency, Activation Scale, etc.

Revisions

A Revision is a point-in-time, immutable snapshot of your code and configuration. Every time you update a KService, Knative automatically generates a new Revision. This makes it incredibly easy to roll back to previous versions or run multiple versions simultaneously for A/B testing.

The Revision view surfaces information like incoming traffic, tags, parent service, concurrency control, etc.

It also provides detailed information about each container, displaying the associated image, exposed ports, and environment variables.

Networking

A clear overview of the ingress settings.

Custom Domains (DomainMapping)

Map your own custom domains to KServices using DomainMapping and ClusterDomainClaim, with an overview of ready status and associated URLs.

Contributors

Thanks to all the contributors who made this release possible!

@kahirokunn, @mudit06mah, @intojhanurag, @Sbragul26, @skoeva, @illume, @joaquimrocha

All Commits

Commit	Author	Message
f0c939b2e	@kahirokunn	knative: Init knative plugin
d0da320d4	@kahirokunn	knative: Add docs and ArtifactHub metadata
e644e5ff2	@kahirokunn	knative: Add cluster selection and install check
92441aab7	@kahirokunn	knative: Define Knative resource classes
661fd0dba	@kahirokunn	knative: Add common UI utilities
00531a596	@kahirokunn	knative: Add KService list and detail views
d2f384b1a	@kahirokunn	knative: Add networking overview
d968d8e36	@kahirokunn	knative: Register routes and sidebar entries
c701e0d2e	@kahirokunn	knative: fix linting issues and apply code formatting
75b32e45f	@kahirokunn	README.md: Add knative to plugin list
9c1e435cf	@kahirokunn	knative: Add permission checks
e0512f596	@kahirokunn	knative: Add copyright headers
b5284dc3e	@kahirokunn	knative: Add missing CRD printer columns
9fdd18e7a	@skoeva	Bump tar to 7.5.3
9734acc9b	@skoeva	Bump lodash to 4.17.23
12ea48a7d	@mudit06mah	knative: Fix package name
4ca0c1eb6	@intojhanurag	knative: utils: time: nullable: Add unit tests
831971d99	@intojhanurag	knative: kservices/List: Memoize the headerProps and actions to reduce renders
ae9388aee	@Sbragul26	frontend: knative: memoize ResourceListView props
2dc52888f	@mudit06mah	knative: kservices: Shift conditions section to common components
b03d60a0a	@mudit06mah	knative: revisions: Add revisions list & detail view, revisions sidebar entry
2e37e92e9	@mudit06mah	knative: common: Refactor condition table
4107ef45d	@mudit06mah	knative: kservice: Refactor kservice sections and add info
65cad08	@mudit06mah	knative: kservice: Polish UI and Add storybooks
dadab55	@mudit06mah	knative: revisions: Polish UI and Add storybooks
d0efb7e	@mudit06mah	knative: networking: Add storybooks
bbdc82b	@mudit06mah	knative: common: Polish UI and add tests
6816b12	@mudit06mah	knative: helpers/config/utils: Add helpers and tests

ai-assistant v0.2.0-alpha

New Changes

Add Holmes agent support

Add MCP support for when ai-assistant is running in desktop app

Other Fixes

1. Add API_KEY for local provider
2. Prevent submit during IME composition

Thanks @aadhil2k4 #495, @minchao #393

flux v0.6.0

New Changes

• 5cae00a flux: Bump version to 0.6.0
• d27538e flux: Update deps with audit fix
• 0551ca9 flux: Persist overview filter preferences in plugin settings
• 0bc6f8b flux: Do not separate charts with space between
• 90a541f flux: Add a way to access plugins settings directly from the Overview
• ad0433d flux: Add a way to show only charts for resources we use
• 10651d5 flux: Fix overview header title size
• 5928802 flux: Reorder sources to show common types first
• b6410f2 flux: Update headlamp-plugin to 0.13.1
• 03fc282 flux: fix api versions
• 1f19d53 flux: add flux crds hierarchy in map feature
• e4339da flux: Count Waiting as processing
• 58aa307 flux: fix namespaced crds fetching for users with limited perissions
• 9679481 flux: kustomization: Fix parse ID for cluster scoped resources
• 2ee3841 flux: overview: Fix error color for failed reconciliations (#493)
• 9734acc Bump lodash to 4.17.23
• c6063f1 Bump glob to 10.5.0
• 9fdd18e Bump tar to 7.5.3
• 7444b80 flux: Update headlamp-plugin to 0.13.0
• c03272a flux: Bump tar to 7.5.2
• 0645a70 flux: Overview: Flux Checks: Add source-watcher to list of controllers
• f46b4a7 flux: Runtime: Add source-watcher to list of controllers

radius v0.1.1

New Plugin

• The Radius plugin provides a way to visualize and manage Radius applications. It builds upon the work of Radius .

Karpenter 0.2.0

New release of the Headlamp plugin for Karpenter.

• Restore missing menu items & set engine compatibility. Thanks to @RachanaVenka #491
• Enable EKS Auto Mode Support. Thanks to @awsfanboy #409
• Upgrade dependencies, and minor typescript fixes. Thanks to @skoeva and @illume
• Updates to README. Thanks to @SinghaAnirban005 #390
• Published Karpenter container images (for 0.1.0 and 0.2.0). Thanks to @illume

See below for some more information, and some screenshots.

Enable EKS Auto Mode Support

This release introduces comprehensive improvements to how the Headlamp Karpenter plugin detects and handles different Karpenter deployment types (EKS Auto Mode, self-installed), enhancing both the user experience and maintainability. The changes include automatic deployment detection, dynamic configuration and UI adaptation, improved error handling, and updated documentation to reflect these enhancements.

EKS Auto Mode

EKS self-installed Karpenter

Restore missing menu items & set engine compatibility

addresses a runtime error causing the Karpenter menu to only display the NodeClass section. It resolves the undefined error on getMainAPIGroup to ensure NodePool, Pending Pods, and Scaling views render correctly.

• Runtime Fix: Patched createNodeClassClass.ts and List views to mock the customResourceDefinition property. This satisfies the Headlamp runtime requirement and prevents getMainAPIGroup failures.
• Compatibility: Added the engines field to package.json to declare support for Headlamp >=0.20.0. This resolves "Incompatible" errors on updated clusters.