gpd-execute-phase

Wave-based parallel execution groups phase plans into execution waves using dependency analysis, coordinating subagent spawning to maximize safe parallel work while maintaining guardrails. It automates plan discovery and groups tasks by dependencies and risk.

Wave-based orchestration is a parallel execution method that groups phase plans into sequential waves based on dependency analysis and risk, coordinating subagents to execute each wave concurrently while enforcing checkpoints and convention locks.

Verification gates enforce first-result, pre-fanout, and review cadences during parallel execution to ensure reproducible results. Checkpoints validate task completion before subsequent waves execute, maintaining reliability across spawned subagents.

Context-aware execution adapts to autonomy and research mode settings by spawning pre-execution agents when needed, adjusting the orchestration flow and reporting progress in real time to match the configured execution context.

Automated subagent spawning coordinates parallelizable work by analyzing plan dependencies and grouping tasks into execution waves. This maximizes concurrent processing while enforcing convention locks and verification gates for safe orchestration.

Use checkpointed phase orchestration when managing complex phase plans with parallelizable dependencies, requiring reproducible results through verification gates, and needing real-time progress reporting across multiple concurrent subagents.