subagent-driven-development

Subagent-driven development executes each task with a fresh agent and mandatory code reviews between tasks, preventing context pollution and catching issues early. This keeps your session continuous while enforcing quality gates at every step.

Fresh agents per task eliminate context interference and confusion in complex plans. Each task gets isolated execution, reducing cognitive load and preventing earlier decisions from biasing later work, while code reviews between tasks catch problems before they compound.

Yes. Dispatch a specialized agent for each component, execute it in isolation, and trigger automated code review after every task completes. This workflow handles multi-component features like the five independent parts of a single feature.

Test-driven development writes tests before implementation, ensuring requirements drive code design. This Skill enforces TDD within each task dispatch and requires code reviews between tasks, catching design and test coverage gaps early rather than late-stage integration.

Use subagent dispatch when your plan contains mostly independent tasks that benefit from isolation, fresh reasoning, and mandatory review cycles. It prevents one task's assumptions from polluting the next and is ideal for feature implementations with discrete, parallel components.

Yes. This Skill scales to handle complex feature plans with multiple independent components. Each component gets a dedicated agent and code review, enabling fast iteration and per-task isolation across all components in a single development branch.