Skip to content
Back to skills

003 Name Skill 1be12f56

BSecurity

TRIZ systematic innovation methodology with AI-enhanced prompts. Use when: (1) Technical contradiction - improve A but B worsens, (2) Physical contradiction - need opposite properties, (3) Cross-industry solutions via FOS/MOS, (4) Technology evolution prediction, (5) Complex engineering problems. Triggers: "TRIZ", "contradiction", "inventive", "trade-off", "improve without worsening", "ข้อขัดแย้งทางเทคนิค", "innovation breakthrough"

  • 9 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added October 11, 2026
developmentgoapiperformance

Works with

  • api

Security analysis

B88/100
  • criticalImpersonates system messages to override safety constraints

Pro shows the line behind each finding and how to fix it

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 003-name-skill_1be12f56 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 003 Name Skill 1be12f56?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for 003 Name Skill 1be12f56
[![Security: B — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-003-name-skill-1be12f56/badge)](https://www.skillsdirectory.com/skills/tools-only-003-name-skill-1be12f56)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

SKILL.md
---
name: triz
description: |
  TRIZ systematic innovation methodology with AI-enhanced prompts. Use when: (1) Technical contradiction - improve A but B worsens, (2) Physical contradiction - need opposite properties, (3) Cross-industry solutions via FOS/MOS, (4) Technology evolution prediction, (5) Complex engineering problems. Triggers: "TRIZ", "contradiction", "inventive", "trade-off", "improve without worsening", "ข้อขัดแย้งทางเทคนิค", "innovation breakthrough"
---

# TRIZ Skill

Systematic innovation via Theory of Inventive Problem Solving. AI-enhanced.

## Problem Routing

| Problem Type | Tool | Reference |
|-------------|------|-----------|
| "Improve A but B worsens" | Contradiction Matrix | [40-principles.md](references/40-principles.md) |
| "Need opposite properties" | Separation Principles | Below |
| "System not working" | Su-Field Analysis | [advanced.md](references/advanced.md) |
| "How will tech evolve?" | Evolution Trends | [advanced.md](references/advanced.md) |
| "What do others do?" | FOS (cross-industry) | [ai-prompts.md](references/ai-prompts.md) |
| "Very complex problem" | ARIZ Algorithm | [ai-prompts.md](references/ai-prompts.md) |

## 6-Step Process

```
1. DEFINE IFR    → "The [system] ITSELF [does X] WITHOUT [cost/harm]"
2. IDENTIFY      → What contradiction? (Technical or Physical)
3. MAP           → Which of 39 parameters? [39-parameters.md]
4. RETRIEVE      → Matrix suggests which principles?
5. GENERATE      → Apply each principle specifically
6. EVALUATE      → Feasibility? Implementation?
```

## Step 1: Ideal Final Result (IFR)

> **"The [object] ITSELF [performs function] WITHOUT [cost/harm/complexity]"**

Formula: `Ideality = Benefits / (Cost + Harm)`

Examples:
- "The pipe itself prevents leaks" (not: add sensors)
- "The code itself fixes bugs" (not: add more tests)

## Step 2: Identify Contradiction

**Technical:** Improving A worsens B
```
"If we [improve A], then [B gets worse]"
→ ถ้าเราทำให้รถเร็วขึ้น, ประสิทธิภาพน้ำมันแย่ลง
```

**Physical:** Same element needs opposite properties
```
"[Element] must be [Property] for X AND [Opposite] for Y"
→ API ต้อง complex (power users) AND simple (beginners)
```

## Step 3: Map to 39 Parameters

See [39-parameters.md](references/39-parameters.md). Common ones:

| # | Parameter | Software Equivalent |
|---|-----------|---------------------|
| 9 | Speed | Performance, latency |
| 27 | Reliability | Uptime, MTBF |
| 33 | Ease of operation | UX, usability |
| 36 | Complexity | Code complexity |
| 39 | Productivity | Throughput |

## Step 4: Top 10 Principles

| # | Principle | Modern Example |
|---|-----------|----------------|
| 1 | Segmentation | Microservices |
| 2 | Taking Out | Separation of concerns |
| 10 | Preliminary Action | Caching |
| 13 | The Other Way Round | Event-driven vs polling |
| 15 | Dynamics | Adaptive algorithms |
| 24 | Intermediary | Middleware, adapters |
| 25 | Self-Service | Self-healing systems |
| 35 | Parameter Changes | Transform data format |

Full list: [40-principles.md](references/40-principles.md)

## Step 5: Physical Contradiction → Separation

| Separation | Strategy | Example |
|------------|----------|---------|
| **In Time** | Different times | Landing gear: extend/retract |
| **In Space** | Different locations | Pencil: hard core, soft eraser |
| **In Condition** | Different conditions | Smart glass: transparent/opaque |
| **In Scale** | Different levels | Water: liquid macro, molecules nano |

## Creative Mode: FOS/MOS

**Function Oriented Search (FOS):** Find how OTHER industries solve same function.

```
1. ABSTRACT → "Remove ice" → "Separate materials"
2. SEARCH → Find 5+ industries with similar function
3. TRANSFER → Adapt mechanism to your problem
```

**Method Oriented Search (MOS):** Apply known method to NEW domains.

See [ai-prompts.md](references/ai-prompts.md) for detailed prompts.

## Output Format

```markdown
## Problem: [Restated]

## IFR: "The [system] itself [does X] without [cost/harm]"

## Contradiction:
- Type: Technical / Physical
- Improving: Parameter #__
- Worsening: Parameter #__

## Principles: [#, #, #]

## Solutions:
### Principle #X: [Name]
- Application: [How]
- Idea: [Concrete solution]
- Feasibility: High/Medium/Low

## Next Steps:
1. [Prototype which solution]
2. [Validation approach]
```

## References

| Type | File | Content |
|------|------|---------|
| Core | [40-principles.md](references/40-principles.md) | All 40 principles + examples |
| Core | [39-parameters.md](references/39-parameters.md) | All 39 parameters |
| Advanced | [advanced.md](references/advanced.md) | Su-Field, 76 Standards, ARIZ, Evolution |
| AI | [ai-prompts.md](references/ai-prompts.md) | Ready-to-use prompt templates |
| AI | [methodology.md](references/methodology.md) | TRIZ + LLM integration |
| Examples | [examples.md](references/examples.md) | Case studies (Samsung, SpaceX, Netflix) |

## Related Skills

- `/generate-creative-ideas` — Complement with broader brainstorming
- `/deep-research` — Research cross-industry solutions (FOS/MOS)
- `/boost-intel` — Evaluate trade-offs systematically
- `/problem-solving` — Structure the problem before applying TRIZ

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.