Six Thinking Hats Analysis

Perform structured multi-perspective analysis using Edward de Bono's Six Thinking Hats technique. This skill launches parallel agents to explore different viewpoints simultaneously.

Workflow

Step 1: Launch Parallel Analysis Agents

Spin up 5 parallel agents to analyze simultaneously:

Agent 1 - White Hat (Facts):

code

Analyze "[TOPIC]" using WHITE HAT thinking:

Use web search to find all known facts, data, statistics about this topic. Note any information gaps.

Format as a bullet list of facts with source and a separate section for information gaps.

Agent 2 - Red Hat (Emotions):

code

Analyze "[TOPIC]" using RED HAT thinking:

What are the immediate emotional reactions to this? What feels right or wrong intuitively?

Format as a list of gut reactions and intuitions.

Agent 3 - Black Hat (Caution):

code

Analyze "[TOPIC]" using BLACK HAT thinking:

Identify all risks, downsides, obstacles, and failure modes. What could go wrong?

Format as a bullet list of concerns and potential problems.

Agent 4 - Yellow Hat (Optimism):

code

Analyze "[TOPIC]" using YELLOW HAT thinking:

Highlight advantages, benefits, ROI, and strategic fit. Why could this succeed?

Format as a bullet list of opportunities and advantages.

Agent 5 - Green Hat (Creativity):

code

Analyze "[TOPIC]" using GREEN HAT thinking:

Generate at least 5 novel ideas, unconventional approaches, or alternatives not yet considered. Think outside the box.

Format as a numbered list with brief explanations.

Step 2: Blue Hat Synthesis

After all agents complete, perform the Blue Hat synthesis yourself:

•Consolidate insights from all perspectives
•Identify patterns and contradictions
•Rank findings by relevance
•Recommend concrete next actions

Output Format

Present the final analysis using this structure:

markdown

## Analysis: [Topic]

**Key Insights:** [2-3 main takeaways]
**Tensions:** [Any conflicting perspectives]
**Recommended Actions:**
1. [Action 1]
2. [Action 2]
3. [Action 3]

### Facts
[Key facts and data gaps]

### Emotions
[Gut reactions and intuitions]

### Risks
[Cautions and potential problems]

### Benefits
[Advantages and opportunities]

### Ideas
[Creative alternatives and novel approaches]

Agent Configuration

When launching Task agents:

•Use subagent_type: "generalPurpose" for each analysis agent
•Set readonly: true since these are analysis-only tasks

Tips

•Be specific with the topic to reduce ambiguity
•Iterate if any hat's output seems incomplete
•Human review the Blue Hat synthesis for final decisions
•For complex topics, consider running multiple passes