Platform Lens / Our direction
Built for the AI era.
Platform Lens connects the Kubernetes workspace engineers use today with the AI workflows they are building next. Lumen brings live operational evidence and AI integration together, with Crossplane UI built in and MCP, agentic development, routing and GitOps integrations on the roadmap.
01 / The foundation
Useful today. Ready for what comes next.
Lumen starts with the live environment: resources, relationships, events, logs and configuration. AI becomes more useful when that operational context stays connected to the engineer’s workflow.
AI integration in the workspace
Ask questions about resources and logs using supported cloud or local model providers. Bring relevant Kubernetes evidence into the conversation, compare explanations with live state and keep engineering judgement close to the answer.
ExploreCrossplane UI, built in
Bring Crossplane into the Kubernetes workspace. Understand the platform abstractions your teams use alongside the resources and operational context behind them, without treating infrastructure as a separate investigation.
Explore02 / Connected engineering
Context, tools and change. In the same loop.
The same engineering question should stay connected as it moves from a resource view to an AI conversation and, eventually, a proposed change. Our direction keeps the environment, evidence and review visible throughout that workflow.
- 01
Live platform context
Cluster and namespace, resource relationships, logs, events and configuration provide the evidence.
- 02
Models & approved tools
AI integration, MCP interoperability and planned routing connect the workflow to useful capabilities.
- 03
Reviewed changes
Diffs, validation and Git-based review keep the intended change and its owner visible.
03 / Coming soon
The next layer of the workspace.
These are our next product priorities. The detail below describes the workflows we’re building toward, rather than a fixed release schedule.
MCP support
Give AI workflows a structured way to request relevant platform evidence through MCP. Our focus is useful resource context, clear tool capabilities and scoped connections, so an agent can work with the environment rather than rely on copied fragments.
ExploreAgentic development
Move from a question to a proposed engineering change. Our focus is task planning, context-aware YAML and configuration changes, validation, reviewable diffs and a deliberate handoff into your delivery process.
ExploreAgent & model router setup
Bring gateway and router configuration into view, including tools such as agentgateway. The direction covers provider and model selection, tool endpoints, routing and fallback policies, access boundaries and request visibility.
ExploreFlux support
Connect source revisions and reconciliation state with live resources. Planned workflows focus on Kustomizations, HelmReleases and source objects, helping engineers follow stalled reconciliation and understand what GitOps intends to deploy.
ExploreArgo CD support
Bring application sync and health into the investigation. Our direction includes desired-versus-live differences, resource relationships and deployment evidence, so a failed sync or unhealthy application is easier to trace back to its cause.
ExploreOperational visibility for AI
Extend investigations with model and tool request traces, latency, usage and failure signals. We’re exploring how engineers can follow an agent workflow across its dependencies and connect failures to platform evidence.
Explore04 / GitOps direction
Keep the source of truth in view.
A controller can report a successful reconciliation while a workload still needs investigation. Upcoming Flux and Argo CD support will connect desired configuration and source revision with the resources running in the cluster, so engineers can inspect both delivery state and application behaviour.
- Follow the relationship between a source revision, controller state and affected workload.
- Separate sync or reconciliation status from application health.
- Review differences and ownership before proposing a change to a controller-managed resource.
SOURCE REVISION → DESIRED CONFIGURATION
RECONCILIATION → LIVE RESOURCES
DIFF & REVIEW → REPOSITORY CHANGE
05 / How we build
More capability. Clear engineering ownership.
AI-ready platforms need useful interfaces and dependable operating boundaries. These principles guide our agentic development and routing work.
- 01
Context before action
Make the target environment and relevant evidence clear. Keep context collection deliberate, especially when logs or resource content may contain sensitive data.
- 02
Explicit connections & permissions
Make model providers, MCP endpoints and tool access understandable. Use scoped credentials and keep consequential actions behind the appropriate review.
- 03
Changes you can inspect
Prefer visible plans, diffs and validation results. Work toward repository-based proposals where GitOps owns the desired state.
- 04
Evidence after delivery
Follow the resulting workload and controller state. Use errors, traces and operational feedback to improve the next iteration.
Product direction / Useful answers
What’s built in. What’s coming.
Is Crossplane UI already built in?
Yes. Lumen includes a built-in Crossplane UI. Deeper composition authoring, reusable templates and validation workflows are a separate area of future development.
Does AI integration mean autonomous production changes?
Current AI integration supports context-aware investigation with configured providers. Agentic development is coming soon and will focus on reviewable engineering work, with permissions and approval appropriate to the environment.
How does MCP fit into Platform Lens?
MCP support is coming soon. It provides an interoperability path for tools and context; the workspace direction keeps connections, credentials and available tool actions explicit.
Will Flux and Argo CD replace existing GitOps workflows?
The planned integrations bring GitOps context into Lumen. Flux and Argo CD remain responsible for their reconciliation workflows, and your repositories remain the source of desired state.
When will these features be available?
Coming soon identifies our next priorities, rather than a committed delivery date. Releases will describe supported versions, setup requirements and available workflows as features ship.
Where do fleet operations and composition authoring fit?
Shared fleet visibility and deeper composition authoring remain part of our longer-term direction. We’re building outward from the Kubernetes IDE, AI integration and Crossplane UI rather than making an optional operator a prerequisite for core desktop workflows.
Help shape the next release
Bring your platform workflow.
Tell us how your team uses AI, Crossplane and GitOps today. Real engineering workflows help us prioritise the next integrations and make them useful in practice.