| 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:
- Ask the tool's core question
- Produce a concrete, specific answer (not generic platitudes)
- 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:
- 관찰+추상화 (Observe + Abstract): What is the hidden essence beneath the surface?
- 패턴+유추 (Pattern + Analogy): What structurally identical problem in another field already has a solution?
- 감정이입+몸 (Empathize + Body Think): Become the problem's subject. What do you feel? What does your gut say?
- 모형+놀이 (Model + Play): Build a minimal model with no constraints. What happens when you run it?
- 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 | | |
