Artificial intelligence agent orchestration defines how one or more agents receive work, exchange context, use tools and pass results between workflow stages. The orchestrator may choose agents and models, enforce dependencies, track state and decide when human input is required.
A development workflow might route an issue through research, implementation, validation and review agents. Effective orchestration keeps those transitions observable and preserves the original intent so that automation remains understandable and correctable.
