What problem does it solve? Getting genuinely diverse perspectives from a single AI is hard — one model tends to voice every side the same way and drift toward easy agreement. This Skill runs a roundtable where multiple BMAD agents or custom personas argue, clash, and build on each other like real people, surfacing angles no single voice would reach. ## Core Features & Use Cases - Multi-persona roundtables: Run lively group discussions between installed BMAD agents or custom personas, with four execution modes (session, auto, subagent, agent-team) controlling whether one mind voices everyone or real independent agents think separately. - Custom party authoring: Create themed casts, one-off personas, open-cast rooms, or distill customer data into AI focus-group panels, saved as TOML overrides via the bmad-customize flow. - Per-party memory: Each remembered party keeps an append-only memlog so grudges, alliances, callbacks, and outcomes carry across sessions. - Use Case: Run the built-in Code Review Crew — a security engineer, an adversary, an edge-case hunter, a craftsman, and a pragmatist — in subagent mode so each lens reviews your code independently before they argue about what actually matters. ## Quick Start Ask the AI to start party mode so the installed agents can debate your topic, or request a specific room such as the Code Review Crew to review your latest changes.
