deep-thinking

When This Skill Activates

• User says "think harder", "ultrathink", "think step by step", "think deeply"
• Complex architecture or system design questions
• Debugging that requires root cause analysis
• Trade-off analysis between multiple approaches
• Code review requiring security/performance deep dive
• Any problem with multiple valid solutions

Extended Reasoning Process

Phase 1: Problem Decomposition

Before answering, break the problem into components:

1. What is the core question being asked?
2. What are the constraints (performance, security, maintainability)?
3. What context do I need that I don't have?
4. What assumptions am I making?

Phase 2: Multi-Path Analysis

For each viable approach:

1. Describe the approach in 2-3 sentences
2. Pros: What makes this approach good?
3. Cons: What are the downsides or risks?
4. When to use: Under what conditions is this best?

Phase 3: Edge Case Hunting

Actively look for:

• What happens with empty/null inputs?
• What happens at scale (1M+ records)?
• What happens with concurrent access?
• What happens when dependencies fail?
• What are the security implications?

Phase 4: Recommendation

Provide a clear recommendation with:

1. The chosen approach and why
2. Key implementation steps
3. What to watch out for
4. How to validate it works

Self-Critique Loop

After initial analysis, critique your own thinking:

• Did I consider all stakeholders?
• Am I biased toward familiar solutions?
• What would a senior engineer challenge here?
• Is there a simpler solution I overlooked?

Output Format

## Analysis: [Problem Title]

### Understanding
[Restate the problem to confirm understanding]

### Approaches Considered
1. **[Approach A]**: [Brief description]
   - Pros: ...
   - Cons: ...

2. **[Approach B]**: [Brief description]
   - Pros: ...
   - Cons: ...

### Edge Cases & Risks
- [Edge case 1]: [How to handle]
- [Edge case 2]: [How to handle]

### Recommendation
[Clear recommendation with reasoning]

### Implementation Steps
1. [Step 1]
2. [Step 2]
...

### Validation
[How to verify the solution works]

Key Principle

Take the time to think thoroughly. A well-reasoned 2-minute response beats a rushed 10-second response that misses critical issues.