What problem does it solve?
This Skill provides a safe, repeatable way to delegate work to isolated helper agents without mixing their context into the main conversation, making parallel work, validation, and long-running tasks easier to manage.
Core Features & Use Cases
- Agent delegation: Launch dedicated sub agents for exploration, implementation, review, or verification.
- Lifecycle control: Track each run from spawn to watch to cleanup so background processes do not linger.
- Safe coordination: Use a strict run ID and tmux-based workflow to avoid collisions, orphaned sessions, and accidental cleanup of unrelated agents.
- Use Case: Split a complex coding task into several independent subtasks, let each helper agent work in isolation, then merge the results after monitoring completion.
Quick Start
Use the subagent skill to start a dedicated helper agent, monitor its progress, and explicitly clean it up when the task is finished.