agent-planner

Task decomposition converts a feature brief into a phased, ordered implementation plan with clear dependencies, per-step success criteria, and recommended agent composition. It breaks work into independently executable steps mapped for sequential or parallel execution.

Dependency mapping identifies required task ordering and highlights parallel execution opportunities to optimize delivery. It marks which implementation steps must wait for previous outputs and which can run concurrently within the phased plan.

Task breakdown decomposes a product roadmap into phased steps with recommended agent types and input/output expectations for each step. This aligns multi-agent execution by defining exactly what each agent must produce to satisfy downstream dependencies.

Automated task planning generates per-step 'done when' success criteria and respects operational constraints like max turns and timeouts. You receive actionable outputs specifying expected inputs, outputs, and agent recommendations for orchestration.

Engineering roadmaps with tight timelines benefit from phase decomposition that identifies major phases like research, design, implement, test, and deploy. It produces a 6 to 12 step implementation plan optimized for available team size and delivery constraints.

Aligning multi-agent execution requires generating step-by-step tasks with agent recommendations and dependency mapping. This ensures each execution agent receives defined inputs and produces expected outputs matching the downstream step requirements.