Adversarial Team Brainstorm

Use when a question needs structured-contention exploration across multiple parallel founding / design / analysis lenses with cross-fire critique, not a single-agent monologue. Runs a reusable multi-team tournament: diverge (N teams generate independently) → cross-fire (teams attack each other) → refine (teams revise) → arbiter (synthesizes winners, attaches kill criteria). Reusable by forge (design exploration), alf (review cycles), agent-teams (tournament mode), founder-ideation, and any caller that needs adversarially-validated outputs with explicit kill criteria. Trigger on: "generate N ideas for X", "tournament", "cross-fire review", "four teams on this problem", "adversarial brainstorm", "kill-criteria ranked outputs".

joogy06 Updated 5 repo stars

File contents

joogy06/agent-foundry commit 559f2afe3e

Frequently asked questions

npx get-skill add joogy06/adversarial-team-brainstorm