task-breakdown-subagent

Task decomposition for subagent execution is achieved by generating a standardized JSON artifact that defines individual implementation tasks, dependencies, and references to requirements for cross-review processes. This structured format ensures clarity and traceability in multi-agent workflows.

Breaking down a feature specification is best handled by creating a Task Breakdown JSON artifact from the attached document, which a dedicated Reviewer role then uses to perform quality checks before development begins.

Cross-review in task breakdown works through a dedicated Reviewer role that evaluates the generated JSON artifact, managing feedback loops with an iterative BLOCK-iteration limit to drive efficient problem-solving and refinement.

Yes, task decomposition supports managing feedback loops using an iterative BLOCK-iteration limit, allowing the Reviewer role to cycle through quality checks and refinement until the task breakdown JSON meets the required standards.

Yes, you need an attached specification document to use as input. The Skill parses this document to define specific task types, dependencies, and references to requirements, outputting a structured JSON artifact for subagent execution.

Structured JSON artifacts for task decomposition should be used when breaking down large tasks in multi-agent or specialized role workflows, ensuring clarity and traceability while enabling cross-review processes between an Architect and a QA Engineer.