RedTeam

Deploys 32 parallel adversarial agents to stress-test arguments and synthesize counter-arguments.

Updated Jul 10, 2026
One-click install
npx skills add https://github.com/Kaleb-Rupe/aurora --skill redteam-kaleb-rupe
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: RedTeam
Source: https://github.com/Kaleb-Rupe/aurora/tree/main/claude/skills/Thinking/RedTeam
Command: npx skills add https://github.com/Kaleb-Rupe/aurora --skill redteam-kaleb-rupe

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Arguments, plans, and proposals often contain hidden assumptions and fatal flaws that a single review pass misses. This Skill subjects content to structured adversarial analysis from 32 expert perspectives, surfacing the weaknesses that could collapse an entire plan before you commit to it. ## Core Features & Use Cases - Parallel Analysis: Decomposes an argument into 24 atomic claims, deploys 32 agents (engineers, architects, pentesters, interns) to examine strengths and weaknesses, then produces an 8-point steelman and an 8-point counter-argument. - Adversarial Validation: Runs a three-round battle-of-bots protocol — competing proposals, brutal critique, collaborative synthesis — to produce a solution stronger than any single draft. - Use Case: Before committing to a microservices migration, ask for a red team review of the plan. The Skill returns the strongest version of the proposal alongside the most devastating counter-argument, exposing the one assumption that could sink the migration. ## Quick Start Red team this architecture proposal and give me the steelman plus the strongest counter-argument.

Frequently Asked Questions about RedTeam

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I red team an argument or plan with AI agents?

Provide the argument or plan and request a red team analysis. The Skill decomposes it into 24 atomic claims, launches 32 parallel agents with distinct adversarial personas, then synthesizes their findings into a steelman and an 8-point counter-argument.

What is the difference between ParallelAnalysis and AdversarialValidation?

ParallelAnalysis stress-tests existing content by attacking it from 32 perspectives to find fatal flaws. AdversarialValidation produces new content through a three-round competition of proposals, brutal critique, and collaborative synthesis.

When should I use adversarial validation instead of a normal review?

Use it for architecture decisions, feature specs, and content where quality matters more than speed. Avoid it for simple decisions with obvious answers, time-critical situations, or problems where expert consensus already exists.

How long does a full 32-agent red team analysis take?

The documented estimate is roughly 9 to 15 minutes total: 2-3 minutes for decomposition, 3-5 for parallel analysis, 2-3 for synthesis, and 2-4 for the steelman and counter-argument phases.

What are the limitations of parallel adversarial analysis?

Results degrade if proposals lack genuine diversity, the critique is too gentle or too destructive, or the synthesis merely blends rather than improves. It also depends on the FirstPrinciples skill for decomposition and constraint classification.