What problem does it solve?
Manually executing multi-task implementation plans in coding agent sessions leads to context pollution, inconsistent task adherence, and missed spec requirements, wasting time on rework and manual per-task review.
Core Features & Use Cases
- Isolated Subagent Execution: Dispatches a fresh, focused subagent per independent task to eliminate context bloat and keep implementation aligned with task-specific requirements.
- Automated Review Gates: Runs spec compliance and code quality reviews after each task, with iterative fix loops to resolve issues before proceeding to the next task.
- Use Case: Ideal for software developers using coding agents (like Claude Code, Codex) to execute well-specified implementation plans with independent tasks, ensuring consistent quality and strict plan adherence without manual overhead for each step.
Quick Start
Use the subagent-driven-development skill to execute the approved implementation plan for the new user profile feature, dispatching a subagent per independent task and reviewing each for spec compliance before moving to the next.