orchestrate

You can coordinate multiple agents for parallel task processing by defining an agents-library.json file and invoking the orchestrator with a task description. It selects up to four specialized agents, launches them concurrently, and synthesizes their outputs into a final report.

Multi-agent orchestration is the concurrent execution of specialized agents to tackle multi-faceted tasks. You need it when a project benefits from simultaneous research, implementation, analysis, and verification across domains like coding, design, or business planning.

To start automated task orchestration, provide a task description using the /orchestrate command. The system loads your agents-library.json file, matches task keywords to select up to four agents, executes them in parallel, and synthesizes a structured executive summary.

Yes, multi-agent orchestration works for software development and strategic planning. It applies to any multi-faceted task requiring concurrent research, implementation, analysis, and verification, synthesizing results across diverse domains into a structured executive summary.

A key limitation of parallel agent orchestration is the maximum capacity of four agents per task. Additionally, it requires a predefined agents-library.json file in the same directory to function, limiting ad-hoc agent creation and selection.