Platform UI

The KubeOpera dashboard is where you manage clusters, ship applications, watch health and cost, respond to incidents and work with AI agents. This page gives you a map of it. Each area has its own page in this section.

Two experiences, one dashboard

What you see depends on who you are:

• Platform administrators operate the infrastructure: clusters, node pools, tenants, users and platform settings.
• Tenant users work inside their own organization: their applications, CloudSpaces, pipelines, monitoring and AI operations.

The menus are genuinely different, not a trimmed copy of each other — each is designed around the job its users do.

Platform administrator navigation

Dashboards
├── Dashboard
├── Cost
├── Performance
├── Security
├── SLOs
├── APM
├── Traces
├── AI Optimizer
└── Actions
Resources
├── Workloads
├── Cloud Spaces
├── Environments
└── AI Agents
Clusters
├── All Clusters
└── Multi-Cluster
Pipelines
Build Configs
Analytics
Incidents
AI Agents
├── Overview
├── Agent Runs          ← watch agents reason live
├── Observability
├── Analysis
├── Actions
└── Recommendations
Node Pools
Settings
├── Users
├── Groups
├── Permissions
├── Tenants
└── AI Provider Key
Support
├── Pending Tickets
├── Create a Ticket
└── Live Chat

Tenant navigation

Tenants don't see infrastructure they don't own — there is no Clusters, Node Pools or platform Settings section.

Dashboard
Resources
├── Workloads
├── Cloud Spaces
├── Environments
└── Pipelines
Monitoring
├── Cost
├── Performance
├── Security
├── Incidents
├── SLOs
├── APM
└── Traces
AI Operations
├── AI Agents
├── AI Runs
├── Observability
├── Analysis
└── Recommendations
Team Members
Support
├── Tickets
└── Live Chats

Team Members lives at the tenant level rather than inside a CloudSpace: a CloudSpace is an infrastructure boundary, while membership belongs to the organization as a whole.

Where to go next in this section

Page	What you'll learn
Dashboard	Read your cluster's health, utilization and observability at a glance.
Clusters	Register, connect or provision clusters and compare your fleet.
Pipelines	Follow CI/CD runs stage by stage.
Analytics	Forecasts, anomalies and scaling approvals.
Incidents	Triage and resolve incidents with runbooks and timelines.
AI Agents	Launch and watch reasoning agents.
App Creation Flow	Deploy an application from source or image.
Multi-Tenancy	How tenants are isolated and managed.
CloudSpaces	Give a team its own virtual cluster.

How the dashboard talks to the backend

The browser never calls backend services directly. Every request goes to a Next.js API route under /api/*, which runs on the server and:

• authenticates the caller with their session token;
• only forwards to known service hostnames (an allowlist that blocks server-side request forgery);
• tries the configured service URL first, then the in-cluster service name;
• returns JSON with a consistent error shape;
• supports regular requests and Server-Sent Events for live streams.

Dashboard route	Service
/api/kubeopera/*	kubeopera-api
/api/k8s-monitor/*	k8s-monitor
/api/security/*	security-api
/api/cicd/*	cicd-gateway
/api/anomalies/*	anomaly-detector
/api/forecasts/*	predictive-scaler
/api/incidents/*	incident-manager
/api/agents/observability/*	observability-agent-srv
/api/agents/analysis/*	analysis-agent-srv
/api/agents/action/*	action-agent-srv
/api/agents/feedback/*	feedback-agent-srv
/api/agents/recommendation/*	recommendation-agent-srv
/api/agents/runtime/*	agent-runtime
/api/ai/chat	Claude (streaming)

Each service's URL is set with an environment variable such as KUBEOPERA_API_BASE_URL. See Configuration for the full list and the Service port reference for defaults.

AI Chat

An AI assistant is available from every page. Ask it questions in plain language and it answers from live data, calling KubeOpera's tools as it goes — you can see each tool call as it happens. It can:

• report cluster health, cost and security posture;
• find anomalies, forecasts and incidents;
• read the AI pipeline's telemetry, analysis, actions and recommendations;
• check the AI Optimizer's status and reports.

The chat respects your role. Tenant users get tools scoped to their own organization and applications — for example, their apps' profiles and AI advice — and never see other tenants' data.

Try asking:

• "Why did the health score of production drop in the last hour?"
• "Which namespaces cost the most this month, and what can I right-size?"
• "Summarize open incidents and what's been done about them."