task-decomposition

Skill: Task Decomposition

Load trigger: "load task-decomposition skill" > DORA: Cap 3 (AI-Accessible Internal Data) Token cost: Low

Purpose

Break the specification and design into clear, actionable tasks and subtasks.

Responsibilities

• Identify major work units from spec and design
• Break work into atomic, buildable subtasks
• Ensure each task maps to acceptance criteria
• Produce a structured task list

Inputs

• specification.md
• design.md

Outputs

• tasks.json

Decomposition Rules

Atomicity

• Each task is independently completable
• Each task produces a verifiable output
• Each task can be implemented in a single PR (≤ 400 lines)
• No task depends on work from multiple other tasks simultaneously

Completeness

• All spec requirements covered by tasks
• All design decisions reflected in tasks
• No orphaned requirements (requirement → task mapping clear)
• Edge cases and error handling included as tasks

Clarity

• Each task has a single, unambiguous goal
• Acceptance criteria are testable
• File paths and components are named explicitly
• Constraints and out-of-scope items documented

Task Format

{
  "task_id": "TASK-NNN",
  "title": "Short imperative title",
  "description": "What this task accomplishes",
  "acceptance_criteria": ["Specific, testable assertion"],
  "context_files": ["path/to/file — reason"],
  "estimated_lines": 150,
  "assigned_agent": "build | test | docs | security",
  "depends_on": ["TASK-NNN"],
  "dora_capability": "Cap N — Name"
}

Success Criteria

• Tasks are complete, atomic, and unambiguous
• Every spec requirement maps to at least one task
• No task exceeds 400 changed lines