CLI vs dashboard
When to use nr versus the NodeRings UI for provider agent operations
The dashboard and the CLI drive the same provider APIs. Choose based on who is operating and whether the work must be repeatable.
Decision guide
| Task | Prefer UI | Prefer CLI |
|---|---|---|
| First-time organization / domain verify | Yes | |
| Explore agents and copy an install command | Yes | Also fine |
| Bootstrap k3s + Liqo + peering on a VM | Install command or CLI | nr cluster register |
| CI / GitOps / headless VM | Yes | |
| Mint & rotate service account tokens | Yes | nr sa … |
| Day-2 verify / debug on the agent host | nr cluster verify / debug | |
| Create platforms, nodes, plans, IPAM | Yes (primary) | Via Provider API if automating |
| Marketplace preflight / live validation | Yes |
Two paths to the same agent
Dashboard: create agent → copy install command → run on VM
CLI: nr cluster register (creates agent + installs)
Both end at: peered agent (+ optional operator) ready for platforms/nodesUsing --agent-id with nr cluster register bridges the paths: create the agent in the UI, then let the CLI perform the host-side install without calling CreateAgent again.
What the CLI does not replace
The CLI is strongest at agent host bootstrap and diagnostics. It does not replace the full provider console for selling capacity: platforms, nodes, VM templates, IPAM, plans, preflight, and billing still live in the dashboard (or the Provider API).
Start with Get started (Proxmox), Get started (VirtFusion), or Get started (SolusVM 2) for the full marketplace path, and use this CLI section whenever the work is “make this Ubuntu VM into a peered agent.”