discuss

To moderate multi-agent discussions, this Skill orchestrates structured AI agent teams with predefined personas, enabling human observers to participate and steer conversations while automatically managing turn-taking and convergence evaluation.

Structured AI agent collaboration involves generating diverse discussant personas with distinct viewpoints to conduct ideation sessions, design critiques, and policy debates, automatically evaluating convergence to synthesize and end discussions.

Yes, you can use Claude Code environments to host multi-agent debates. The Skill performs pre-flight environment validation and team checks to ensure discussion readiness before initializing the moderated session.

To start a moderated discussion, define the agent roles, topic, and optional user participation. The Skill automatically configures agent personas, initializes the session, and handles transcription with synthesis.

Convergence evaluation in AI agent discussions automatically assesses the transcript to determine when the debate has reached synthesis, ending the session once the diverse viewpoints have been thoroughly explored and aligned.