What problem does it solve? Coordinating multiple coding agents and terminal processes across workspaces, tabs, and panes is error-prone when done manually. This Skill lets an agent inspect and control a Herdr session programmatically, so it can split panes, start agents, send prompts, and read output without guessing IDs or disrupting the user's focus. ## Core Features & Use Cases - Session Discovery: List workspaces, tabs, panes, and live agents, reading stable IDs from JSON responses instead of predicting them. - Agent Orchestration: Start recognized coding agents in available shell panes, submit prompts with lifecycle-aware waits, and send keystrokes to interactive agent UIs. - Command Execution: Run ordinary commands in sibling panes, wait for matching output, and read transcripts using viewport, recent, or unwrapped sources. - Use Case: While working in a Herdr pane, ask the agent to spin up a reviewer agent in a right-hand split, prompt it to review the current diff, and report back actionable findings. ## Quick Start Ask the agent to verify it is running inside Herdr and then start a reviewer agent in a new pane to the right to review the current diff.