subagent-development

Executing an implementation plan with independent subagents involves dispatching a fresh subagent for each task within your session, ensuring focused development. This approach prevents context pollution by isolating task execution and enforcing a two-stage review process for spec compliance and code quality.

The two-stage review process for iterative development mandates a sequential evaluation of each completed task. It first checks the code for strict spec compliance against the implementation plan, then evaluates the code quality, ensuring high standards before merging the subagent's output.

Preventing context pollution when handling multiple development tasks requires assigning each task to a new, dedicated subagent. This task decomposition strategy ensures each subagent maintains a focused context, systematically building components without interference from other plan segments.

Subagent task execution requires integration with specific development workflow skills for a complete process. You need to combine it with tools for writing-plans, git-worktrees, test-driven-development, and finishing-development to manage plans, isolate code, verify tests, and finalize sessions.

Breaking down a feature plan for focused code review requires task decomposition into discrete, manageable units. By assigning each unit to an independent subagent, you ensure the two-stage review process evaluates spec compliance and code quality for each component systematically.

A subagent development workflow requires a dedicated subagent for each task to guarantee focused development and rigorous review. This isolation prevents context bleed between tasks, ensuring each component is built and reviewed systematically for higher quality iterative development.