agent-teams

Spawning parallel agents for design critique involves launching multiple subagents with distinct perspectives like Minimalist, Bold, or Playful on the same task. You collect all generated variations and synthesize them into a final unified result to reduce blind spots.

Multi-perspective synthesis in prompt engineering is the process of spawning parallel agents with distinct viewpoints to generate diverse outputs, then collecting and synthesizing those variations into a superior final result to reduce bias and surface insights.

To synthesize outputs from multiple perspectives, the Skill collects all variations generated by the parallel subagents and applies a final unification step. This process merges diverse viewpoints into a superior final result, reducing blind spots and improving quality.

Yes, you can use parallel agents for complex problem solving without dependencies. The Skill supports spawning subagents per perspective and synthesizing results directly, requiring no external components or prerequisites to execute the teamwork workflow.

The default perspectives for spawning parallel agents are Minimalist, Bold, Technical, Playful, and Elegant. These distinct viewpoints allow the subagents to generate diverse variations for tasks like brainstorming, design critique, and policy drafting.