subagent-driven-development

Read the plan once, extract all tasks into a todo list, then dispatch a fresh implementer subagent per task with the full task text in context. After each implementation, run a spec compliance review followed by a code quality review before marking the task complete.

Two-stage review means checking spec compliance first, then code quality. The spec reviewer verifies the implementation matches the original task requirements with no scope creep; only after it passes does the quality reviewer check style, error handling, tests, and security.

Use it when you have a written implementation plan with mostly independent tasks and quality matters. Fresh subagents prevent context pollution from accumulated state, and automated reviews catch issues early, at the cost of more subagent invocations per task.

Each task should represent 2-5 minutes of focused work, such as creating a single model or endpoint. Broad tasks like implementing an entire authentication system should be split into smaller units like password hashing, token generation, and registration endpoints.

The implementer subagent, or a new fix subagent, addresses the issues, then the reviewer re-reviews until approved. You should never proceed to the next task with open critical or important issues, and never let the implementer's self-review replace the actual review stages.