gsd-execute-phase

Wave-based parallel execution manages task dependencies by grouping plans into parallel waves, ensuring correct execution order while processing multiple tasks across subagents simultaneously to reduce overall completion time.

The best way to execute large-scale tasks in parallel phases is by orchestrating task execution across multiple subagents using wave-based parallelism, which manages pacing, quota allocation, and dependency groupings for efficient workflow automation.

Yes, you can limit parallel task execution to a specific subset by setting a wave number with the '--wave N' flag, focusing the phase execution on a targeted group of tasks within the overall workflow.

Yes, phase execution requires the ability to spawn subagents and handle task contexts independently, as the workflow relies on orchestrating multiple parallel processes across these subagents to manage dependencies and quota.

You should use staged rollout for workflow automation when managing large-scale task execution that requires parallel processing, pacing control, and quota management to ensure workflow integrity across dependent task groupings.