Overview

A project groups resources inside a tenant. Where a tenant is a hard boundary with its own quota, a project is an organisational convenience with its own consumption view.
Navigation: Settings → Projects. Titled Manage Projects, subtitled “View and manage project resources and members”.

Cards, not a table

Projects display as cards rather than rows, because each one carries a consumption breakdown. There’s no search, no filters, and no toolbar — just the cards and Add Project.

What a card shows

Along the top: the project name, a clickable member chip with admin and user counts, a state badge, a Manage Resources button, and a delete button. Then five resource tiles, each with a percentage, a bar, {used} / {total}, and how much remains: Bars change colour as they fill — amber at 75%, red at 90%. A footer summarises Running VMs, Stopped VMs, IPs, Volumes, and the created date.
An uncapped resource shows 0% rather than “unlimited”, because the percentage is calculated against a limit of zero. A 0% tile doesn’t necessarily mean nothing is used — check the {used} / {total} figures beside it.

Creating a project

Add Project opens Create A Project — “Create a new project to organize your cloud resources and team members.”

What isn’t finished

Project management is partially built, and the gaps matter for how you use it:Manage Resources is a placeholder. The button opens a dialog that lists CPU, Memory, and Storage read-only and says: “Resource management for projects is coming soon. You’ll be able to adjust CPU, memory, and storage allocations here.” You cannot change a project’s allocations. The cost figures shown in that dialog are hardcoded zeros, not real pricing.You cannot move existing resources into a project. There’s no assignment action anywhere.The practical consequence: set the project when you create a resource, because you can’t reorganise afterwards. Treat project membership as a decision made at creation time.
A “plan management” feature exists in the code but has no button to trigger it, so there’s nothing to find. Ignore any reference to project plans.

Projects versus tenants

Use a project when the same team wants resources labelled and tracked separately — staging versus production, or per-client work. Use a tenant when one group genuinely must not see or consume another’s.

Worked example: tracking spend per client

An agency running infrastructure for three clients in one tenant.
1

Create a project per client, before creating any resources

client-acme, client-globex, client-initech. This has to come first — resources can’t be moved in later.
2

Assign each resource at creation time

Every machine, volume, and network for a client goes into that client’s project as you create it.
3

Read consumption per client from the cards

Each card’s CPU, Memory, Storage, VMs, and Networks tiles give you that client’s footprint without manual tallying.
4

Use the footer for a quick efficiency check

Running VMs versus Stopped VMs shows what’s actually working versus what’s idle but still billed. Stopped machines still consume storage and network quota — see Quotas and limits.
5

Reach for tenants if clients must be isolated

If a client requires that their resources are invisible to others, projects aren’t enough. Move to sub-tenants.
Deleting a project is not a way to clean up resources — and conversely, don’t assume it’s harmless. Verify what belongs to a project before deleting it, since there’s no assignment view to check afterwards.

Empty and error states

Tenants

Hard boundaries with their own quotas.

Quotas and limits

What the tiles are measured against.

Settings

The rest of the Settings area.