gsd-execute-phase

Wave-based parallel execution groups plans by dependency analysis and spawns subagents to run them concurrently. It orchestrates plan discovery and collects results to verify phase completion.

Wave-based parallel execution is an orchestration method that groups interdependent plans into sequential waves, spawning subagents for concurrent processing within each wave while verifying overall phase completion.

Yes, you can run a single wave using the optional --wave flag filter. This allows targeted execution of specific wave numbers while still supporting robust result collection and error handling.

Subagent orchestration handles errors and retries during plan execution by verifying phase completion after waves finish. It applies context budgeting and robust result collection to manage interdependent plan failures.

The best way to orchestrate phase plans is using wave-based parallel execution, which applies dependency analysis to group plans, spawns subagents concurrently, and verifies phase completion while handling errors and retries.

Yes, parallel workflow execution supports context budgeting for subagent orchestration. It satisfies requirements for flag handling and robust result collection while executing interdependent plans across multiple waves.