What problem does it solve? Managing multiple parallel coding agents across git worktrees is chaotic without a central dispatcher. This Skill turns the AI into a coordinator that spawns, monitors, instructs, and merges worktree agents using the workmux CLI, so parallel tasks stay organized and conflicts are avoided. ## Core Features & Use Cases - Agent Spawning: Write self-contained prompt files and launch background agents with workmux add, each in its own git worktree and branch. - Monitoring & Communication: Check agent statuses, wait for completion with timeouts, capture output, and send follow-up instructions or skill commands like /merge. - Fan-out / Fan-in Workflows: Spawn several agents in parallel, wait for all to finish, review captured output, then merge branches one at a time to avoid conflicts. - Use Case: You need an auth module, API tests, and docs updated simultaneously. The coordinator writes three prompt files, spawns three worktree agents, waits for completion, reviews each output, and sequentially triggers merges. ## Quick Start Ask the AI to coordinate worktree agents by spawning a background agent for each task, waiting for them to finish, and merging their branches one at a time.