critical-thinking

name	critical-thinking
description	Analyze arguments, detect biases, evaluate claims, and improve reasoning. Use when asked to fact-check, identify logical fallacies, evaluate arguments, analyze predictions, find root causes, or think adversarially about plans. Triggers include "evaluate this argument", "logical fallacies", "fact check", "analyze the claims", "identify biases", "devil's advocate", "red team this", "root cause".

Critical Thinking

Rigorous analysis of arguments, claims, and reasoning.

Pattern Selection

Intent	Pattern	When to Use
Claim evaluation	analyze_claims	Assess truth claims with evidence
Prediction analysis	extract_predictions	Identify and assess predictions
Extraordinary claims	extract_extraordinary_claims	Claims contradicting consensus
Controversial ideas	extract_controversial_ideas	Contested viewpoints analysis
Error analysis	analyze_mistakes	Learn from past errors
Problem finding	extract_primary_problem	Root cause identification
Solution analysis	extract_primary_solution	Evaluate proposed solutions
Adversarial thinking	t_red_team_thinking	Find weaknesses in plans
Decision upgrade	create_upgrade_pack	Improve decision-making
Novel insights	extract_alpha	Most surprising/novel ideas
Thought organization	create_idea_compass	Structure complex ideas
Mind mapping	create_markmap_visualization	Visual thinking maps

Decision Flow

User request
    │
    ├─ "evaluate claims/fact check" ──→ analyze_claims
    ├─ "predictions/forecasts" ──→ extract_predictions
    ├─ "controversial/contested" ──→ extract_controversial_ideas
    ├─ "what went wrong/mistakes" ──→ analyze_mistakes
    ├─ "root cause/core problem" ──→ extract_primary_problem
    ├─ "red team/devil's advocate" ──→ t_red_team_thinking
    ├─ "surprising/novel insights" ──→ extract_alpha
    └─ "organize my thinking" ──→ create_idea_compass

Pattern References

See references/ for full patterns:

• analyze_claims.md
• extract_predictions.md
• analyze_mistakes.md
• extract_primary_problem.md
• t_red_team_thinking.md
• extract_alpha.md
• create_idea_compass.md

Output Guidelines

• Always distinguish claims from evidence
• Rate confidence levels explicitly
• Identify logical fallacies by name
• Present steelman versions of arguments before critiquing
• Note when evidence is insufficient to conclude
• Acknowledge uncertainty and areas of genuine disagreement

Chaining Suggestions

• After analyze_claims → offer extract_primary_problem to find root issues
• After t_red_team_thinking → offer create_upgrade_pack to address weaknesses
• After extract_primary_problem → offer extract_primary_solution for solutions