task-decomposition

Task decomposition breaks down high-level objectives into atomic, executable tasks that can be independently tracked and verified across iterations. It generates an ordered task graph with explicit preconditions and dependencies to guide multi-step execution.

You can plan cross-file changes by generating an ordered, atomic task plan with explicit dependencies. This workflow enforces task-level granularity, ensuring each task targets at most two to three files for small, testable changes during multi-step work.

Persistent task tracking works by creating and updating task_plan.md and progress.md files. These planning artifacts maintain the state of your atomic tasks and dependencies across context windows, enabling continuous collaborative planning.

The best way to ensure task-level granularity is to enforce a limit where each atomic task targets at most two to three files. This approach guarantees small, testable changes and produces a repeatable workflow for teams.

Yes, task decomposition is versatile for breaking down large features, multi-file changes, and cross-team work. It handles cross-file changes and multi-step work by generating explicit preconditions and an ordered task graph for any software project scope.

You should avoid atomic task decomposition for simple, single-file modifications that do not require explicit preconditions or persistent progress tracking. It is designed for complex, multi-step work and cross-file changes that benefit from a dependency-aware ordered task graph.