subagent-driven-development

Execute complex implementation plans consistently by delegating focused tasks to specialized subagents. This approach breaks down plans into independent tasks, dispatching fresh subagents with complete context for each, ensuring systematic validation and reliable code delivery instead of relying on a single coding pass.

Validate spec compliance and code quality using a two-stage review process. First, check spec compliance to catch missing requirements, then perform code quality reviews to identify engineering issues, ensuring your feature development meets all specifications before final integration.

Break down a multi-step software feature into independent implementation tasks using task decomposition. This workflow dispatches fresh subagents with complete context for each focused task, enabling independent implementation and review cycles before integration verification.

Yes, subagent delegation applies to refactoring tasks requiring independent review cycles. It orchestrates systematic validation by dispatching fresh subagents for each refactoring task, verifying spec compliance and code quality before completing the integration verification process.

Single-pass code generation fails for multi-step engineering features because it lacks systematic validation. Without delegated subagents and structured review loops, a single coding pass misses missing requirements, overlooks engineering issues, and skips the integration verification needed for reliable changes.

Avoid subagent-driven workflows for simple, single-step software implementation tasks that do not require independent review cycles. If your task needs no task decomposition, spec compliance checks, or multi-stage code quality reviews, a single coding pass is more efficient than orchestrating delegated subagents.