bmad-party-mode

To get multiple AI agent perspectives, this Skill orchestrates a multi-agent roundtable by spawning independent subagents for each voice. Each agent thinks for itself instead of being roleplayed by one model, providing diverse expert perspectives for project decisions.

Starting a multi-agent roundtable discussion requires asking the Skill to bring in the appropriate BMAD agents. It resolves the agent roster, loads project context, configures model selection, and spawns independent subagents for your software planning conversation.

Yes, multi-agent orchestration supports configurable model selection. The Skill provides flexible orchestration that includes solo mode, agent roster resolution, and project-context loading to accommodate different discussion depths and model preferences.

You need multi-agent orchestration when a single shallow AI answer is insufficient for complex project tradeoffs. It is ideal for generating product, architecture, and delivery discussions simultaneously, or when you need one agent to respond directly to another agent's point.

Roleplayed agents suffer from a single model simulating multiple voices, leading to shallow answers. Independent subagents overcome this by spawning separately for each voice, ensuring authentic independent reasoning and direct agent-to-agent reactions during collaborative analysis.