feat: bound container health pane so logs stay visible on large stacks (#1556)

* feat: bound container health pane so logs stay visible on large stacks

On the stack overview page, multi-container stacks (8+ services) pushed the
logs pane off-screen because the Command Center card had shrink-0 and no
scroll mechanism. This reworks the desktop layout for stacks with more than
one container:

- The Command Center card is bounded at max-h-[42%] so it never consumes
  the full left column. The container list scrolls internally via Sencho's
  existing Radix ScrollArea.
- A summary strip shows total / running / paused / unhealthy counts above
  the list, with a Compact / Detailed density toggle. Detailed (default)
  shows CPU / Mem / Net sparklines; Compact hides them for a denser list.
- The logs pane has a guaranteed min-h-[180px] so it is always reachable.
- Single-container stacks are untouched: the card keeps its original
  shrink-0 behaviour, and no summary strip or density toggle appears.

Mobile is unaffected (it already uses segmented Health / Logs / Compose
panes).

* fix: key ContainersHealth by node+stack so density resets on navigation

The density toggle state (compact/detailed) is local to ContainersHealth.
Without a key, selecting Compact on one multi-container stack would persist
when navigating to a single-container stack, hiding sparklines while also
hiding the toggle itself. Keying by node+stack ensures a clean remount.

Also added test coverage for zero containers and density reset on remount.

* fix: remove unused rerender variable from ContainersHealth test

docs/features/stack-management.mdx

@@ -203,7 +203,7 @@ The same links button appears on each container row and on the update cards in [
 
 ## Container health strip
 
-Below the header, each container in the stack gets a single row that answers "is this piece working, and how do I reach it?" without expanding anything.
+Below the header, each container in the stack gets a single row that answers "is this piece working, and how do I reach it?" When the stack has multiple containers, a summary strip appears above the list showing total, running, paused, and unhealthy counts, along with a **Compact / Detailed** toggle. Compact mode shows status, name, uptime, port, and action buttons; detailed mode (the default) adds CPU, memory, and network I/O sparklines.
 
 <Frame>
   <img src="/images/stack-view/containers.png" alt="CONTAINERS section showing a container card with health badge, uptime, port mapping, open link, and CPU, memory, and network stat tiles" />