bmad-party-mode

You orchestrate multi-agent roundtable discussions by spawning independent subagents to provide diverse, expert-driven perspectives on user queries. This facilitates complex decision-making by managing agent rosters, context summaries, and cross-agent interactions.

The best way to simulate a brainstorming session with specialized AI agents is to spawn independent subagents as distinct processes. This ensures authentic, non-convergent viewpoints by maintaining independent agent personas and state during debates.

Multi-agent collaboration prevents echo-chamber thinking by launching multiple independent agents to debate topics from diverse perspectives. This dynamic roster management ensures non-convergent viewpoints rather than relying on a single AI response.

Yes, you need the Agent tool and the project-specific configuration resolver to run multi-agent conversations. These dependencies are required to spawn independent subagents and maintain their distinct personas and state throughout the discussion.

Yes, you can use dynamic roster management to automatically resolve agent configurations and select the most relevant experts for any given topic. This allows specialized agents like a software architect or security expert to debate the best path forward.

You should not use subagent spawning for simple queries that do not require diverse, expert-driven perspectives. The orchestration of independent subagents is designed for complex decision-making and collaborative brainstorming sessions, not straightforward tasks.