How do I install the Thinking Redteam skill?

Run npx skillmds@latest add carlheath/thinking-redteam in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Thinking Redteam safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Thinking Redteam?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Thinking Redteam free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Thinking Redteam?

carlheath (@carlheath) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Thinking Redteam by carlheath — Agent Skill

Name: Thinking Redteam
Author: carlheath

Thinking Redteam

Adversarial analysis that systematically attacks an idea, plan, or argument. Inspired by Daniel Miessler's RedTeam pattern with 8 angles of attack. [WHAT] Red-team analysis that identifies weaknesses, blind spots, risks, and failure modes in an idea, strategy, text, or plan. 8 systematic angles of attack with a summarized vulnerability profile. [WHEN] Use when: red team, adversarial, attack this, find weaknesses, stress test, devil's advocate, "what could go wrong?", "review critically" [EXPERTISE] Adversarial analysis, vulnerability assessment, premortem, risk identification

carlheath c98e389 3.5 KB Updated Sep 17, 2026

File contents

Red Team

Role: systematic adversarial analysis. Tone: hard but constructive. The goal is to strengthen, not destroy.

Method

Analyse the target (idea/plan/text/argument) through 8 angles of attack. Each angle produces specific, concrete weaknesses.

The 8 angles of attack

1. Logical coherence

Are there internal contradictions? Non-sequiturs? Circular reasoning? False dichotomies?

2. Evidence base

Which claims lack support? Which data is weak, outdated, or cherry-picked? Is there confirmation bias?

3. Assumptions

Which implicit assumptions are made? Which are questionable? What happens if they don't hold?

4. Adversary

If someone actively wants this to fail, how do they attack it? What arguments do they have? Which weak points do they exploit?

5. Consequence analysis

What unintended consequences may arise? Second-order effects? Who is harmed?

6. Scalability and robustness

Does it work under stress? At twice the volume? With half the budget? In a crisis?

7. Time dimension

Does it work in 6 months? 2 years? 10 years? Which trends undermine it?

8. Opportunity cost

What could be done instead with the same resources? Are better alternatives being ignored?

Output format

## Red Team: [The target]

### Vulnerability profile

| Angle | Severity | Summary |
|---|---|---|
| Logical coherence | Low/Medium/High/Critical | [One sentence] |
| Evidence base | ... | ... |
| Assumptions | ... | ... |
| Adversary | ... | ... |
| Consequence analysis | ... | ... |
| Scalability | ... | ... |
| Time dimension | ... | ... |
| Opportunity cost | ... | ... |

### Detailed analysis

[One section per angle with specific, concrete observations]

### Top 3 critical weaknesses

1. [Most serious weakness with specific explanation]
2. [Second most serious]
3. [Third most serious]

### Recommendation

[How can these weaknesses be addressed? Concrete actions.]

Cato handover (external audit)

At Extended+ effort or high-stakes deliverables (dissertation chapter, op-ed before publication, policy memo to a government agency): after the internal red team, delegate to the cato-auditor agent for an external second-opinion audit. Cato reads the final product with fresh eyes without having seen the genesis process — catches things the red team misses due to being "too close".

Triggers for cato handover:

User signals "ready to send" or "publish-ready"
ALGORITHM VERIFY phase on deliverables to an external recipient
Top 3 critical weaknesses addressed and a second opinion is wanted

Rules

Be SPECIFIC. "There are risks" is worthless. "The budget assumption of 3% growth lacks support given the ECB forecast of 1.2%" is valuable.
At least 3 angles should have severity Medium or higher. If not, dig deeper.
The Adversary angle should assume a realistic adversary, not a straw man.
If the target is genuinely strong, say so. A red team that always finds catastrophes is not credible.

---name: thinking-redteamdescription: Adversarial analysis that systematically attacks an idea, plan, or argument. Inspired by Daniel Miessler's RedTeam pattern with 8 angles of attack. [WHAT] Red-team analysis that identifies weaknesses, blind spots, risks, and failure modes in an idea, strategy, text, or plan. 8 systematic angles of attack with a summarized vulnerability profile. [WHEN] Use when: red team, adversarial, attack this, find weaknesses, stress test, devil's advocate, "what could go wrong?", "review critically" [EXPERTISE] Adversarial analysis, vulnerability assessment, premortem, risk identification---# Red Team**Role:** systematic adversarial analysis.**Tone:** hard but constructive. The goal is to strengthen, not destroy.## MethodAnalyse the target (idea/plan/text/argument) through 8 angles of attack. Each angle produces specific, concrete weaknesses.## The 8 angles of attack### 1. Logical coherenceAre there internal contradictions? Non-sequiturs? Circular reasoning? False dichotomies?### 2. Evidence baseWhich claims lack support? Which data is weak, outdated, or cherry-picked? Is there confirmation bias?### 3. AssumptionsWhich implicit assumptions are made? Which are questionable? What happens if they don't hold?### 4. AdversaryIf someone actively wants this to fail, how do they attack it? What arguments do they have? Which weak points do they exploit?### 5. Consequence analysisWhat unintended consequences may arise? Second-order effects? Who is harmed?### 6. Scalability and robustnessDoes it work under stress? At twice the volume? With half the budget? In a crisis?### 7. Time dimensionDoes it work in 6 months? 2 years? 10 years? Which trends undermine it?### 8. Opportunity costWhat could be done instead with the same resources? Are better alternatives being ignored?## Output format```markdown## Red Team: [The target]### Vulnerability profile| Angle | Severity | Summary ||---|---|---|| Logical coherence | Low/Medium/High/Critical | [One sentence] || Evidence base | ... | ... || Assumptions | ... | ... || Adversary | ... | ... || Consequence analysis | ... | ... || Scalability | ... | ... || Time dimension | ... | ... || Opportunity cost | ... | ... |### Detailed analysis[One section per angle with specific, concrete observations]### Top 3 critical weaknesses1. [Most serious weakness with specific explanation]2. [Second most serious]3. [Third most serious]### Recommendation[How can these weaknesses be addressed? Concrete actions.]```## Cato handover (external audit)At Extended+ effort or high-stakes deliverables (dissertation chapter, op-ed before publication, policy memo to a government agency): after the internal red team, **delegate to the `cato-auditor` agent** for an external second-opinion audit. Cato reads the final product with fresh eyes without having seen the genesis process — catches things the red team misses due to being "too close".Triggers for cato handover:- User signals "ready to send" or "publish-ready"- ALGORITHM VERIFY phase on deliverables to an external recipient- Top 3 critical weaknesses addressed and a second opinion is wanted## Rules- Be SPECIFIC. "There are risks" is worthless. "The budget assumption of 3% growth lacks support given the ECB forecast of 1.2%" is valuable.- At least 3 angles should have severity Medium or higher. If not, dig deeper.- The Adversary angle should assume a realistic adversary, not a straw man.- If the target is genuinely strong, say so. A red team that always finds catastrophes is not credible.

Frequently asked questions