Token导航 LogoToken导航TokenDH.com
研究检索需要联网github未标认证来源可访问clear审计通过

triztriz 分析

Agent Skill

triz 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

823

周安装

35

GitHub Stars

21

下载量

288
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:triz(triz 分析)
来源仓库:https://github.com/thepexcel/agent-skills
仓库路径:skills/triz
安装命令:
npx skills add https://github.com/thepexcel/agent-skills --skill triz
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/thepexcel/agent-skills --skill triz

简介

TRIZ 分析技能用于查找、检索和筛选相关信息,支持基于关键词和任务场景的线索定位。

  • 适用于 Codex、Claude、Cursor、Gemini CLI 中需要快速获取候选结果的研究类任务。
  • 通过 npx skills add 命令从指定 GitHub 仓库安装,需结合原始 README 核验具体用法。
  • 安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写操作。
  • 该技能归类于研究检索类别,适合创新方法和问题解决相关场景。

SKILL.md

TRIZ Skill

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

Problem Routing

Problem TypeToolReference
"Improve A but B worsens"Contradiction Matrix40-principles.md
"Need opposite properties"Separation PrinciplesBelow
"System not working"Su-Field Analysisadvanced.md
"How will tech evolve?"Evolution Trendsadvanced.md
"What do others do?"FOS (cross-industry)ai-prompts.md
"Very complex problem"ARIZ Algorithmai-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. Common ones:

#ParameterSoftware Equivalent
9SpeedPerformance, latency
27ReliabilityUptime, MTBF
33Ease of operationUX, usability
36ComplexityCode complexity
39ProductivityThroughput

Step 4: Top 10 Principles

#PrincipleModern Example
1SegmentationMicroservices
2Taking OutSeparation of concerns
10Preliminary ActionCaching
13The Other Way RoundEvent-driven vs polling
15DynamicsAdaptive algorithms
24IntermediaryMiddleware, adapters
25Self-ServiceSelf-healing systems
35Parameter ChangesTransform data format

Full list: 40-principles.md

Step 5: Physical Contradiction → Separation

SeparationStrategyExample
In TimeDifferent timesLanding gear: extend/retract
In SpaceDifferent locationsPencil: hard core, soft eraser
In ConditionDifferent conditionsSmart glass: transparent/opaque
In ScaleDifferent levelsWater: 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 for detailed prompts.

Output Format

## 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

TypeFileContent
Core40-principles.mdAll 40 principles + examples
Core39-parameters.mdAll 39 parameters
Advancedadvanced.mdSu-Field, 76 Standards, ARIZ, Evolution
AIai-prompts.mdReady-to-use prompt templates
AImethodology.mdTRIZ + LLM integration
Examplesexamples.mdCase 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

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

30.67%
按下载量换算88

OpenCode

22.48%
按下载量换算65

windsurf

18.46%
按下载量换算53

Codex

14.28%
按下载量换算41

Antigravity

8.93%
按下载量换算26

Gemini CLI

3.37%
按下载量换算10

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

来源信息

继续浏览同类 Skills