What problem does it solve? Coordinating multiple AI coding agents working in parallel git worktrees is error-prone: prompts lack context, merges conflict, and status tracking is manual. This Skill provides a disciplined dispatch workflow for spawning, monitoring, messaging, and merging worktree agents through the workmux CLI. ## Core Features & Use Cases - Parallel Agent Spawning: Write self-contained prompt files first, then launch all agents in background with workmux add -b -P. - Status Monitoring & Waiting: Track agent states (working, waiting, done) with workmux status, wait, and capture, including timeouts and cross-project handle resolution. - Controlled Merging: Send /merge to agents one at a time, waiting between each to avoid branch conflicts. - Use Case: Split a feature into auth module, API tests, and docs tasks; spawn three worktree agents, wait for completion, review captured output, then merge each branch sequentially. ## Quick Start Ask the coordinator to split your feature into independent tasks and run them as parallel worktree agents with workmux, then merge the results one at a time.