bmad-party-mode

Multi-agent orchestration coordinates conversations between installed BMAD agents by selecting participants based on topic relevance and maintaining their individual personalities to facilitate collaborative problem-solving. It processes agent manifests and executes sequential workflows to manage the dialogue.

Dynamic agent selection in a multi-agent conversation evaluates the user's query topic and intelligently matches it to the expertise of available BMAD agents. This ensures only relevant agents participate, maintaining focused and natural group discussions.

Maintaining individual AI agent personalities during group brainstorming is supported by merging each agent's distinct expertise and persona. The orchestration workflow processes the agent manifest to preserve these traits throughout the collaborative session.

Starting a multi-agent brainstorming session requires initiating party mode to activate all available BMAD agents. This orchestrates a group discussion where agents with specializations like marketing, design, and development collaboratively generate ideas.

An agent manifest is required to manage sequential workflows for AI agents because it provides the metadata needed to identify available agents, their personalities, and their expertise for accurate topic relevance matching during orchestration.

Limitations of orchestrating natural conversations across multiple AI agents include the dependency on sequential workflow execution, which may restrict truly parallel real-time interactions, and the requirement for a pre-processed agent manifest to enable dynamic selection.