multi_agent_collaboration

To coordinate multiple agents, first query available agents in the workspace, then initiate cross-agent chats to delegate tasks, pass session context, and execute multi-step workflows requiring diverse expertise.

Multi-agent collaboration is needed when a task requires cross-agent input, context, or workspace data from multiple agents with diverse expertise. It applies to task delegation, information integration, and quality checks across different agents.

To start a cross-agent chat, first query and discover available agents in the workspace, then use the agent chat command to initiate a bidirectional conversation that passes session context when needed to complete the user request.

Yes, multi-agent collaboration supports optional background task execution along with session record management and state tracking, allowing you to monitor progress and integrate information across agents without unnecessary callbacks.

To avoid unnecessary callbacks during multi-agent coordination, enforce explicit safety rules, pass session context only when needed, and manage state through session records rather than triggering redundant cross-agent communication.

Yes, multi-agent collaboration operates without external dependencies, relying entirely on internal agent discovery and chat mechanisms to coordinate diverse expertise and manage multi-step workflows within the workspace.