What problem does it solve? Managing Herdr terminal sessions from an AI agent is error-prone when the agent relies on memorized or copied command syntax that may not match the installed Herdr version. This Skill ensures the agent always loads the live, version-accurate Herdr guidance and correctly identifies the calling pane before issuing control commands. ## Core Features & Use Cases - Live guidance loading: Runs herdr --skill at execution time so command syntax always matches the installed Herdr binary instead of a stale copied reference. - Reliable caller identification: Uses herdr pane current --current to resolve the actual Pi pane making the call, avoiding mistakes caused by UI focus belonging to another session. - Correct process launching: Prefers Herdr operations that supply a command as a pane's initial process rather than racing a second command against shell startup. - Use Case: While working in a Herdr-managed session, ask the agent to open a new tab running a test suite; the agent loads the live Herdr guide, identifies its own pane, and spawns the tab with the command as the initial process. ## Quick Start Ask the agent to use Herdr to open a new pane running your test command in the current workspace.