Agent Hierarchy

Orchestrator v2 topology: DB-first task flow, per-worker dispatch controls, live lane status, and cron fallback.

Auto-refresh every 10s

Orchestration Topology (v2)

You (Owner)

Sets priorities and approves review items

Main Agent (Coordinator)

Owns queue decisions and dispatch commands

Orchestrator v2 (Dispatch Layer)

Backlog 0 · In Progress 0 · Blocked 0

Tasks Store

DB-first persistence with migration-safe fallback

Dispatch Controls

Per-worker controls · Busy 0/3

Live Notifications

Dispatch activity + review updates surfaced in Tasks UI

Alex (Frontend)

idle

Cron fallback: inactive

Riley (Backend)

idle

Cron fallback: inactive

Michael (General/Architecture)

idle

Cron fallback: inactive

Runtime Flow (Orchestrator v2)

Current architecture from intake to worker execution and status notification.

  1. 1. Intake: Owner adds or updates task via Tasks UI.
  2. 2. Store: Task state persists to DB-first tasks store (JSON fallback only during migration).
  3. 3. Dispatch: Orchestrator v2 assigns backlog items per lane and WIP guards.
  4. 4. Execute: Worker lanes (Alex/Riley/Michael) process assigned tasks and return updates.
  5. 5. Notify: Status/dispatch activity is surfaced in Tasks UI and review queue.

Cron Fallback Layer (Resilience)

Active cron workers: 0 / 0

WorkerModeScheduleLast Status
No isolated worker jobs found.