sparks-create

name	sparks-create
description	Apply the 13 thinking tools from "Sparks of Genius" (생각의 탄생) to reframe any problem through creative, cross-disciplinary lenses. Use when conventional analysis falls short, when you need a fresh creative angle, or when the problem spans multiple domains.
user-invocable	true

Sparks — Creative Problem Solving with 13 Thinking Tools

Apply the 13 thinking tools from Robert & Michèle Root-Bernstein's "Sparks of Genius" to break through conventional thinking. While /rethink-unblock targets technical blockers with structured frameworks, /sparks-create targets problems that need creative, cross-disciplinary, and sensory-rich exploration.

Triggers

Use this skill when:

• The problem is not purely technical — it involves design, strategy, communication, or human factors
• Analytical approaches have produced correct but uninspiring solutions
• You need to connect ideas across different domains
• The problem needs a fundamentally new framing, not an incremental fix
• You want to explore a question with creative depth rather than engineering precision

Process

Step 1: Define the Problem

Problem: [What I want to achieve or understand]
Context: [Domain, constraints, what's been tried]
Desired outcome: [What a good solution looks like]

Step 2: Select a Mode

Choose one of three modes based on need:

Mode	Tools applied	When to use
Full (전체)	All 13 tools, sequentially	Deep exploration of a complex, open-ended problem
Quick (단축)	5 combined lenses	Fast creative check, time-boxed brainstorming
Deep (심화)	1-2 tools, thoroughly	Already know which lens is most promising

If the user doesn't specify, default to Quick mode.

Step 3: Apply the Selected Tools

Follow the tool instructions in the relevant section below. For each tool:

1. Ask the tool's core question
2. Produce a concrete, specific answer (not generic platitudes)
3. Bridge to the next tool by connecting insights

Step 4: Synthesize and Recommend

Produce actionable output (see Output Format).

---

The 13 Thinking Tools

1. Observing (관찰)

"보기"가 아니라 "관찰하기" — 피카소의 아버지는 비둘기 발만 반복해서 그리도록 시켰다.

• Describe the problem as if seeing it for the first time, using all senses.
• What is everyone taking for granted that is actually surprising?
• What is the "pigeon's foot" — the one element worth observing repeatedly?

2. Imaging (형상화)

테슬라는 기계를 머릿속에서 조립하고 작동시켰다.

• Visualize the problem as a concrete image. What does it look like?
• Translate to other senses: what does it sound like, feel like, taste like?
• Visualize the solved state. What changed in the image?

3. Abstracting (추상화)

"하나만 제외하고 모든 변수를 제거하라."

• Strip away all details. What is the one-sentence essence?
• What is the simplest possible version of this problem?
• If you could change only one variable, which one matters most?

4. Recognizing Patterns (패턴 인식)

다빈치는 벽의 얼룩에서 풍경을 발견했다.

• What repeats in this problem? What is the underlying rhythm?
• Where has this same pattern appeared in a different time or field?
• Where does the pattern break? That irregularity is a clue.

5. Forming Patterns (패턴 형성)

바흐는 대칭적 패턴 조작으로 독창적 음악을 만들었다.

• Decompose the found pattern into its smallest units.
• Recombine those units in at least 3 new arrangements.
• Which recombination opens the most surprising possibility?

6. Analogizing (유추)

뉴턴: 사과를 끌어당기는 힘이 달까지 뻗친다면?

• Find a functionally identical structure in a completely different domain.
• The analogy must share mechanism, not just surface resemblance.
• What solution already exists in the analogous domain? Reverse-map it.

7. Body Thinking (몸으로 생각하기)

파인먼은 문제를 풀지 않고 "느꼈다."

• If you had to act out this problem physically, what movement would it be?
• Describe the solution process as a physical verb: building, untangling, carving, folding...
• What does your gut say before your mind finishes analyzing?

8. Empathizing (감정이입)

제인 구달은 침팬지 사회에 들어가 그들이 되었다.

• Become the problem's subject (user, system, molecule, customer). What do you feel?
• "If I were this problem itself, why would I exist in this form?"
• What does the world look like from inside the problem?

9. Dimensional Thinking (차원적 사고)

오키프는 꽃을 거대하게 그려서 사람들이 "보게" 만들었다.

• Shift the time scale: what if you had 1 second? 100 years?
• Shift the magnitude: 1 person vs. 1 million? 1 won vs. 1 trillion?
• Flatten it to 2D or expand it to 4D (adding time). What appears?

10. Modeling (모형 만들기)

모형은 직접 경험하기 어려운 것에 접근하는 도구다.

• Build the simplest possible model that captures the core mechanism.
• Run the model. Where does it break? What does the failure reveal?
• What would a napkin sketch or 3-step flowchart of this look like?

11. Playing (놀이)

플레밍: "미생물을 가지고 노는데, 규칙에 익숙해지면 깨뜨리는 게 재미있다."

• Remove all constraints. What would you try just for fun?
• If failure were impossible, what's the most audacious approach?
• Turn the problem into a game. What are the rules? What's the win condition?

12. Transforming (변형)

에드거튼은 사진의 한계를 전자공학으로 풀었다.

• Express the problem in a completely different medium (diagram, story, equation, song).
• Translate the language: technical → everyday, Korean → English, numbers → narrative.
• Borrow a tool from another field. What does it reveal?

13. Synthesizing (통합)

"상상하면서 분석하고, 화가인 동시에 과학자가 되라."

• Which insights from the previous tools resonate with each other?
• Where do logic and intuition point in the same direction?
• State the final answer and trace which tools contributed to it.

---

Quick Mode: 5 Combined Lenses

When using Quick mode, apply these 5 combined lenses instead of all 13 tools:

1. 관찰+추상화 (Observe + Abstract): What is the hidden essence beneath the surface?
2. 패턴+유추 (Pattern + Analogy): What structurally identical problem in another field already has a solution?
3. 감정이입+몸 (Empathize + Body Think): Become the problem's subject. What do you feel? What does your gut say?
4. 모형+놀이 (Model + Play): Build a minimal model with no constraints. What happens when you run it?
5. 변형+통합 (Transform + Synthesize): Re-express the problem in a different medium, then unify all insights.

---

Deep Mode: Single-Tool Drill

When using Deep mode, the user specifies which tool(s) to apply. Go deep on that tool:

• Apply the tool from at least 5 different angles or domains
• Produce 3+ concrete alternatives from that single lens
• Include a "first step" for each alternative

Example: /sparks-create deep analogy — find 5 analogies across nature, music, sports, cooking, and history, then reverse-map the best solution.

---

Output Format

## Sparks: [Problem Title]

**Problem**: [one sentence]
**Context**: [brief background]
**Mode**: [Full / Quick / Deep: tool-name]

### [Tool/Lens Name] Analysis
[Specific insights, 3-5 sentences per tool. Include concrete examples.]

...

### Synthesis

**Key insights** (top 3):
1. [insight + which tool produced it]
2. [insight + which tool produced it]
3. [insight + which tool produced it]

**Recommended approach**: [1-2 sentences]
**First action**: [immediate next step]
**Alternative paths**: [1-2 backup approaches if the first doesn't work]

Anti-Patterns

• Generic creativity theater ("think outside the box!") — every insight must be specific and grounded in the problem
• Skipping the hard tools — body thinking and empathy feel awkward but often produce the strongest breakthroughs
• Decorating an existing answer — the point is to discover NEW directions, not justify what you already planned
• All insight, no action — output must include concrete next steps
• Treating this as a replacement for /rethink-unblock — use /rethink-unblock for technical blockers, /sparks-create for creative reframing

Relationship to Other Thinking Skills

	/rethink-unblock	/sparks-create	/davinci-define
Focus	Technical blockers	Creative reframing	Problem definition & direction setting
Frameworks	4 structured frameworks	13 thinking tools	7 Da Vincian principles
Strength	Engineering precision	Cross-domain imagination	Self-reflection & life compass
Best for	Architecture, debugging, design decisions	Strategy, innovation, multi-domain problems	Career direction, goal setting, life planning
Can combine	Use /rethink-unblock first for technical analysis, then /sparks-create for creative alternatives, then /davinci-define for direction alignment