subagent-driven-development

Automate code review between independent development tasks by dispatching a fresh subagent for each task and enforcing explicit review gates that evaluate implementation quality before proceeding to the next item.

Subagent-driven development is a workflow that dispatches a fresh subagent for each task to maintain clean context boundaries. Use it for plans with three or more independent issues that can be investigated without shared state.

Run parallel execution for independent programming tasks by loading your plan and dispatching subagents simultaneously, ensuring tasks have no shared state so they can execute sequentially or in parallel with quality gates.

Subagent-driven development does not work well for tasks with shared state. It is specifically designed for independent issues or tasks that can be investigated without shared state to maintain clean context boundaries.

You can use subagents to finalize a development branch by performing final development completion checks after all tasks are executed and reviewed, supporting batch planning and branch finalization workflows.

Dispatching a fresh subagent for each task prevents context pollution and enables isolated experimentation. This approach enforces high-quality iteration by performing code review between tasks to catch issues early.