What problem does it solve? When facing multiple implementation tasks, it is hard to decide whether to work solo, spawn parallel agents, run sequential subagents, or use Agent Teams — and choosing wrong wastes tokens or causes merge conflicts. This Skill provides a structured decision process that picks the cheapest strategy satisfying the task's coordination needs. ## Core Features & Use Cases - Four-strategy decision framework: Evaluates solo, parallel agents, sequential subagents, and Agent Teams against task count, independence, file overlap, merge complexity, and communication needs. - Worktree-based parallelism: Uses detached-HEAD git worktrees plus a cherry-pick integration protocol so tasks with trivial file overlap can still run in parallel safely. - Explicit handoff contracts: Generates constraint checklists for downstream execution skills covering isolation strategy, integration ordering, autonomy level, and model selection. - Use Case: Given an implementation plan with 8 tasks, some sharing files, the Skill reads the plan and source files, determines that worktree-isolated parallel agents are viable, and hands off explicit squash-and-cherry-pick instructions to the dispatching skill. ## Quick Start Ask the AI to evaluate the implementation plan and choose the most cost-efficient orchestration strategy for executing these tasks.
