subagent-driven-development

Subagent orchestration automates multi-task plan implementation by dispatching fresh subagents for each independent task and enforcing two-stage reviews. It coordinates implementer, spec reviewer, and code quality reviewer subagents to ensure high-quality code delivery.

To ensure spec compliance during automated code generation, use a two-stage review process that first checks the implementation against the original specification and then assesses overall code quality. This isolates context and prevents contamination across tasks.

You can execute independent development tasks within the same session by using a skill that reads your implementation plan, extracts tasks into TodoWrite entries, and coordinates fresh subagents to complete them sequentially with dual reviews.

Using fresh subagents for each independent task in task orchestration isolates context to avoid contamination. This approach ensures that previous task contexts do not interfere with subsequent implementations, maintaining high code quality and spec fidelity.

Subagent-driven development is designed for executing detailed implementation plans with independent tasks within the same session. It is not intended for plans where tasks have strict dependencies and require sequential context sharing.

Subagent-driven development does not require manual setup for code quality review. It provides full workflow orchestration that automatically coordinates implementer, spec reviewer, and code quality reviewer subagents for each task.