Ref #316, #319
The backend now exposes GET /api/ui/v1/queue/status which returns per-agent concurrency slot usage. This data isn't surfaced in the UI yet.
What to add:
- Per-agent indicator showing running/max/available execution slots
- Visual feedback when an agent is at capacity (all slots occupied)
- Total running count across all agents
API response shape:
{
"enabled": true,
"max_per_agent": 3,
"total_running": 2,
"agents": [{
"agent_node_id": "my-agent",
"running": 2,
"max": 3,
"available": 1
}]
}
In #316 the user had no way to see that jobs were piling up with no concurrency control.