superpowers:subagent-driven-development

You can automate multi-step coding workflows by orchestrating specialized subagents that execute plan tasks in isolated contexts. This approach prevents coordination fatigue and context loss during complex software implementation.

Subagent-driven development is a workflow technique that dispatches isolated implementer and reviewer subagents to enforce test-driven development. It applies automated quality gating to verify spec compliance for each coding task.

To execute a complex software implementation plan, you create a todo list from the plan and dispatch the first implementer subagent. The system automatically delegates subsequent tasks and runs review loops until completion.

Yes, you can use autonomous subagents to enforce code review by applying a strict review loop for every task. The reviewer subagent verifies both spec compliance and code quality before marking any task as complete.

You can track development progress persistently by maintaining a progress ledger throughout the subagent execution. This durable ledger prevents redundant work and ensures recovery from session interruptions.

You should not use subagent-driven development for simple, single-step modifications that lack a formal implementation plan. The orchestration overhead is designed specifically for multi-step coding projects requiring rigorous spec compliance.