← ClaudeAtlas

dialectic-brainstorminglisted

Brainstorm a consequential, no-ground-truth question with a panel of diverse minds, then resolve it dialectically. A casting step derives orthogonal thinking-axes for the question (each a quality to maximize) and casts one persona per axis; the personas propose and cross-critique concrete candidates while a neutral Synthesizer prunes, recombines across axes, and holds a frontier — returning a short list, not one winner. Use for naming (products, tools, projects), positioning, or any choice of consequence with no objectively correct answer where you want many genuinely different perspectives, not one. Triggers: /dialectic-brainstorming, "brainstorm this with a diverse panel", "what should we name…", "give me a shortlist from multiple perspectives".
mehdiforoozandeh/skills · ★ 0 · AI & Automation · score 72
Install: claude install-skill mehdiforoozandeh/skills
# dialectic-brainstorming — diverse minds diverge, dialectic converges A sibling of `dialectic`. Where `dialectic` runs three *structural* roles (thesis/antithesis/ synthesis) to drive one question to a single answer, this skill runs **N diverse persona-minds** that diverge through *different backgrounds*, then a neutral Synthesizer converges them to a **short list** via dialectical sublation. Diversity is the engine of divergence; dialectic is the engine of convergence. Use it for consequential questions with **no ground-truth answer** — naming, positioning, "which of these directions", any judgment call where you want many genuinely different minds rather than one voice averaging over itself. ## Mechanism in one line Cast orthogonal quality-axes → each persona maximizes its quality with concrete candidates → they cross-critique → the Synthesizer prunes weak candidates *using the critiques* and **recombines across axes** (crossover) onto a frontier → return the surviving short list. ## Two modes — pick before casting Each subagent spawn costs ~27k tokens **flat** (the agent's system prompt + tool schemas dominate; the brainstorming text is noise next to it). So **spawn count is the cost**, and the mode is a cost/independence trade: - **Batched (DEFAULT — low-to-mid-stakes).** A *single* subagent role-plays the whole panel in one call: give it all N axes/personas and have it return each persona's `PROPOSE`+`CRITIQUE` **independently and divergently** (instruct it e