What problem does it solve? Executing a multi-task implementation plan in one AI session leads to context pollution, skipped reviews, and lost progress after compaction. This Skill orchestrates plan execution by delegating each task to a fresh subagent with isolated context, enforcing review gates, and tracking progress in a persistent ledger. ## Core Features & Use Cases - Per-task subagent dispatch: Each plan task gets a fresh implementer subagent with a file-based task brief, keeping the controller's context clean for coordination. - Two-stage review gates: Every task diff is reviewed for spec compliance and code quality, followed by a whole-branch final review on the most capable model. - Bounded fix loop with escalation: Up to five fix rounds per task, resuming the original implementer for rounds 1-3 and escalating to a more capable model for rounds 4-5, with adjudicated rulings recorded in a ledger. - Use Case: You have a written implementation plan with eight independent tasks. This Skill creates a git worktree, dispatches implementer and reviewer subagents per task, logs progress to a ledger that survives context compaction, and finishes with a whole-branch review before merge. ## Quick Start Use the subagent-driven-development skill to execute the implementation plan at docs/plans/feature-plan.md task by task with reviews.
