Wall
Wall is the ambient fleet display for Colony Cloud. It has two entry points depending on how you want to use it.
| Entry point | Route | When to use |
|---|---|---|
| In-shell view | /dashboard/wall | Previewing the wall inside the normal dashboard UI |
| Standalone kiosk | /wall | Casting to an office screen, TV, or a second monitor for passive fleet monitoring |
In-shell view
Section titled “In-shell view”Route: /dashboard/wall
The in-shell view renders the same wall display inside the AppShell, so it sits inside your normal navigation chrome. This is where you go to preview the wall or orient yourself before putting it on a shared display.
An Enter wall mode button in the page header opens the standalone kiosk at /wall in the same browser tab, entering fullscreen automatically.
Standalone kiosk
Section titled “Standalone kiosk”Route: /wall
The kiosk is a full-screen, chrome-free view intended for a shared office display or TV. It is rendered outside the AppShell — no sidebar, no header, no navigation.
Fullscreen behaviour
Section titled “Fullscreen behaviour”When the kiosk loads, it automatically requests browser fullscreen. If the browser blocks the request (for example, because it was not triggered by a direct user gesture), the page continues to render normally without fullscreen.
To exit the kiosk, press Escape or exit browser fullscreen by any other means. Either action navigates back to /dashboard/wall in the in-shell view.
Data connection and polling fallback
Section titled “Data connection and polling fallback”The kiosk wraps its own PipelineStateProvider. Because it is rendered outside the AppShell — which normally provides the shared SSE context — useOptionalSSEContext() returns null inside the provider. As a result, the kiosk always falls back to 3-second HTTP polling for pipeline state. There is no SSE stream in kiosk mode.
The staleness indicator dims the display and shows a warning when pipeline data has not updated in more than 15 seconds, whether due to polling delay or a network issue.
What the Wall shows
Section titled “What the Wall shows”The wall display is composed of three sections.
KPI row
Section titled “KPI row”A horizontal strip of six fleet-wide metrics at the top of the display.
| KPI | What it measures |
|---|---|
| Active | All issues whose state is not done — every in-progress, blocked, paused, and rework state is counted. |
| In flight | Issues in the in-review or merge-pending states — work that has reached the review/merge stage of the pipeline. |
| Blocked | Issues that need intervention before the pipeline can advance — needs-clarification, dependency-blocked, failure-blocked, changes-requested, planning, waiting-for-subtasks, paused, and human-review-ready states. |
| Merged today | Issues that transitioned to done state since midnight UTC today. |
| Cost today | LLM token spend by Colony agents since midnight UTC today, in USD. |
| Pending tasks | Pipeline tasks queued but not yet assigned to a Worker. |
State groups
Section titled “State groups”One tile per active pipeline state, showing every issue currently in that state. Each issue tile shows its issue number, repository, and an aging bucket indicating how long the issue has been in its current state. Older issues are highlighted progressively to draw attention to stalled work.
Recent transitions
Section titled “Recent transitions”A scrolling strip at the bottom of the display showing the most recent pipeline state transitions across the fleet — useful for confirming that work is moving.
Staleness indicator
Section titled “Staleness indicator”The display dims and shows a warning banner when the pipeline data has not refreshed in more than 15 seconds. This can occur when the kiosk loses network connectivity or when the polling request is slow to respond.
Relationship to Live
Section titled “Relationship to Live”Both Wall and Live show fleet-wide pipeline activity, but they are designed for different contexts.
| Live | Wall | |
|---|---|---|
| Route | /dashboard/pipeline-live | /dashboard/wall (in-shell) or /wall (kiosk) |
| Intended use | Interactive operator monitoring | Passive ambient display on a shared screen |
| Navigation chrome | Full AppShell with sidebar | Kiosk: none; in-shell: normal AppShell |
| Data connection | SSE stream with HTTP-poll fallback | In-shell: SSE via AppShell; kiosk: always 3 s HTTP poll |
| Interaction | Drill-down into issues, agent activity | Read-only display, no click targets |
Use Live when you are actively monitoring the fleet and need to drill into individual issues or agent activity. Use Wall when you want a persistent, no-interaction overview on a dedicated screen.
For the full tour of dashboards and monitoring surfaces, see Dashboards and Live.