What problem does it solve? Coordinating multiple AI agents on a shared task requires structured messaging, task dispatch, and escalation handling that ad-hoc prompting cannot provide. This Skill routes those coordination needs through the Orca runtime so agents can exchange threaded messages, wait on worker results, and decompose work into task DAGs. ## Core Features & Use Cases - Structured Multi-Agent Coordination: Supports threaded messages, blocking ask/reply flows, task dispatch, worker_done and escalation waits, task DAGs, decision gates, and coordinator loops. - Version-Matched Guidance: Loads the full orchestration reference directly from the installed Orca binary via ORCA skills get orchestration, so instructions never drift from the binary that runs the commands. - Safe CLI Resolution: Selects the correct Orca executable per environment (ORCA_CLI_COMMAND, orca-dev, orca-ide, or orca) and avoids accidentally launching the GNOME Orca screen reader. - Use Case: When a user asks to split a large feature across several agents with a coordinator monitoring progress and gating decisions, this Skill directs the agent to bootstrap Orca, load the current guide, and dispatch the task DAG. ## Quick Start Ask the agent to coordinate multiple workers on a task using Orca orchestration, and it will resolve the correct CLI, run ORCA skills get orchestration to load the full guide, then dispatch and monitor the tasks.
