task-decomposition_skill

Overview

This skill breaks down complex objectives into atomic, actionable goals with explicit dependencies and measurable success criteria. It produces a hierarchical plan that supports parallelization, prioritization, and clear handoffs between agents. Use it to convert high-level requests into execution-ready task lists.

How this skill works

I extract the primary objective, implicit requirements, constraints, and success criteria, then create a top-down goal hierarchy of 3–7 major components. Each component is decomposed into atomic tasks that include inputs, outputs, preconditions, and quality standards. I map dependency types (sequential, parallel, converging, resource) and assign priorities and estimates to create an execution-ready plan.

When to use it

• Complex user requests with multiple moving parts
• Multi-phase projects that require coordination across teams or agents
• When you want to enable parallel execution safely
• Planning features that need database, API, and testing workstreams
• Refactoring or architectural changes with many dependencies

Best practices

• Start with a clear problem statement and success metrics
• Keep tasks atomic: single action, clear inputs/outputs, testable
• Identify and document dependencies explicitly before assigning work
• Include testing and documentation tasks in every decomposition
• Prioritize by blocking impact, user value, and risk reduction

Example use cases

• Design and implement a new API feature with database migrations, business logic, and docs
• Coordinate multiple agents to refactor a storage layer across backends
• Break a bugfix request into reproduce→diagnose→fix→verify→prevent-regression steps
• Plan a phased rollout: research → foundation → core implementation → integration → optimization
• Create a parallelizable test plan and CI tasks for a feature release