What problem does it solve? Coordinating human-first task execution across DAG-structured work slices is error-prone when agents guess at CLI flags, confuse Tasker WorkGroups with Projectman sprints, or reimplement scheduling logic that the hosted domain service already enforces atomically. ## Core Features & Use Cases - Ownership routing: Directs planning and review to Projectman, human-first execution tracking to Tasker, run/lease control to Runner, and durable memory to Agentspace. - DAG-aware status and planning: Reads ready/waiting/blocked/stalled scheduling state and plans slice dependencies with deterministic topological waves via aops loop parallel-plan. - Atomic worker leases: Previews and applies worker claim/release operations with idempotency keys, letting the domain service enforce readiness, terminal gates, and maxParallelSlices. - Use Case: An agent resuming a multi-slice run checks aops runner status to find the next-ready slice, claims it atomically with a worker lease, and releases the lease with the correct terminal state when finished. ## Quick Start Ask the agent to use the aops-cli-tasker skill to check Runner status for your run ID and claim the next ready slice with an atomic worker lease.