What problem does it solve? Coordinating multiple coding agents, shells, and long-running commands across terminals is chaotic without a structured control surface. This Skill lets an agent inspect and control Herdr workspaces, tabs, and panes through the herdr CLI, so it can spawn agents, run commands, read output, and wait for state changes programmatically. ## Core Features & Use Cases - Pane and workspace control: List, split, rename, and organize panes, tabs, and workspaces using stable public IDs parsed from JSON responses. - Agent lifecycle management: Start interactive agents (Codex, Claude Code, pi, OpenCode, OMP) in sibling panes, submit prompts, and wait on agent_status transitions like idle, working, blocked, or done. - Command execution and output reading: Run ordinary commands in other panes, wait for matching output, and read scrollback via visible, recent, recent-unwrapped, or detection sources. - Use Case: While working on a feature, ask your agent to open a reviewer agent in a side pane, have it review the current diff, and report actionable findings without leaving your session. ## Quick Start Use the herdr skill to split my current pane and start a Codex agent that reviews the current diff.