What problem does it solve? Executing a multi-task implementation plan in a single context leads to context pollution, inconsistent quality, and skipped reviews. This Skill orchestrates plan execution by dispatching a fresh subagent for each independent task, enforcing a two-stage review (spec compliance, then code quality) before any task is marked complete. ## Core Features & Use Cases - Isolated Task Execution: Each task runs in a fresh subagent with precisely curated context, preventing cross-task contamination and preserving the controller's coordination context. - Two-Stage Review Gates: Every task passes a spec-compliance review (verifying nothing missing or extra) followed by a code-quality review, with fix-and-re-review loops until both pass. - Status-Based Handling: Implementer subagents report DONE, DONE_WITH_CONCERNS, NEEDS_CONTEXT, or BLOCKED, and the controller responds with targeted actions such as re-dispatching with a stronger model or splitting oversized tasks. - Use Case: You have a written plan with five independent tasks. The controller extracts all task text upfront, dispatches an implementer subagent per task, runs both review stages after each, and finishes with a final whole-implementation code review. ## Quick Start Use subagent-driven development to execute the implementation plan in docs/plans/feature-plan.md, dispatching a fresh subagent for each task with spec and quality reviews.
