parallel-planning

Wave-based execution structures tasks into sequential phases where multiple tasks run concurrently when safe. This approach organizes work by dependencies, allowing tasks with no conflicts to execute in parallel within each wave, reducing overall delivery time for multi-module projects and feature rollouts.

Parallel planning detects file, database, and external resource conflicts before execution to avoid race conditions. The skill enforces dependency awareness, ensuring tasks respect resource constraints and only execute when safe, preventing conflicts that would compromise data integrity.

Yes, wave-based execution applies to data pipelines, multi-module projects, and feature rollouts where tasks can run concurrently. The skill generates constrained execution plans that respect dependencies and integrates with the goop_wave_execute API to drive deterministic rollout.

Wave planning builds on task decomposition by organizing work into sequential phases where independent tasks execute in parallel. This adds coordination layer and conflict detection to simple decomposition, enabling faster delivery through structured parallelism rather than sequential execution.

Define the feature with its tasks and dependencies, then use the goop_wave_execute API to generate a wave plan that groups tasks by execution phase. The API produces a constrained schedule respecting dependencies, which you then execute deterministically across waves.

Conflict detection identifies when tasks compete for the same file, database, or external resource. The skill assigns conflicting tasks to different waves to prevent race conditions, ensuring dependencies are satisfied and resources are accessed safely across parallel execution.