Token导航 LogoToken导航

Agent Skill · 研究检索

made-to-stick(粘起来)

made-to-stick 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。

made-to-stick 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 4,671。

总安装 30,576周安装 1,293GitHub Stars 779下载 10,712核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:made-to-stick(粘起来)
来源仓库:https://github.com/wondelai/skills
仓库路径:skills/made-to-stick
安装命令:
npx skills add https://github.com/wondelai/skills --skill made-to-stick
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。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/wondelai/skills --skill made-to-stick

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

粘起来用于查找、检索和筛选相关信息。适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词、任务场景或来源线索快速定位候选结果时使用。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。注意该工具当前无底部简介,功能以实际仓库内容为准。

来源仓库内路径:skills/made-to-stick

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/wondelai/skills --skill made-to-stick;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

made-to-stick 怎么安装?

在终端执行 npx skills add https://github.com/wondelai/skills --skill made-to-stick。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

made-to-stick 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

made-to-stick 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

made-to-stick 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

made-to-stick 用的人多吗?

公开数据是累计安装 30,576 次,近一周安装 1,293 次,下载 10,712 次,GitHub 779 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

made-to-stick 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
made-to-stick

粘起来 · 本页

30,576779只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Made to Stick Framework

A framework for crafting ideas and messages that are understood, remembered, and have lasting impact. Based on decades of research into why some ideas survive and others die.

Core Principle

The Curse of Knowledge is the single greatest barrier to effective communication. Once we know something, we can't imagine not knowing it. This makes us bad at explaining our ideas to others.

The foundation: Sticky ideas aren't born — they're made. The SUCCESs framework provides six principles that make any idea more memorable and impactful.

Scoring

Goal: 10/10. When reviewing or creating messaging (copy, presentations, campaigns, onboarding), rate 0-10 based on SUCCESs principles. A 10/10 means the message is simple, surprising, concrete, credible, emotional, and wrapped in a story; lower scores indicate forgettable communication. Always provide current score and improvements to reach 10/10.

The SUCCESs Framework

Six principles that make ideas stick:

S - Simple
U - Unexpected
C - Concrete
C - Credible
E - Emotional
S - Stories

Not a checklist — a toolkit. Not every sticky idea uses all six. But the stickiest ideas tend to use most of them.

1. Simple

Core concept: Find the core of the idea and share it compactly.

Simple ≠ dumbed down. Simple means finding the essential core and expressing it in a compact way. It means ruthless prioritization.

The Commander's Intent:

  • Military term: If everything goes wrong, what ONE thing must we accomplish?
  • For messaging: If people remember ONE thing about your product, what should it be?

The inverted pyramid:

  • Lead with the most important thing
  • Add detail in order of decreasing importance
  • Readers who stop anywhere still got the core

Techniques for simplicity:

TechniqueHow It WorksExample
Core messageStrip to the essentialSouthwest: "THE low-fare airline"
AnalogyExplain new via known"It's like Uber for dog walking"
GenerativeCore idea that generates behavior"Names, names, names" (local newspaper motto)
PrioritizeForce-rank what matters"If you say 3 things, you say nothing"

Application to product messaging:

Before (Complex)After (Simple)
"AI-powered, cloud-native customer engagement platform with omnichannel capabilities""Talk to all your customers in one place"
"We leverage machine learning algorithms to optimize conversion funnels""We find why visitors don't buy and fix it"
"Enterprise-grade project management with Gantt charts, resource allocation...""The simplest way to manage projects"

The test: Can you explain it to a smart 12-year-old? If not, simplify.

Warning: Don't oversimplify to the point of meaninglessness. "We make the world better" is simple but empty.

See: references/simple.md for simplification exercises and templates.

2. Unexpected

Core concept: Get attention by breaking patterns. Hold attention by creating curiosity gaps.

Two tasks:

  1. Get attention → Surprise (violate expectations)
  2. Hold attention → Interest (create curiosity gaps)

Surprise:

  • Identify the core message
  • Figure out the counterintuitive implication
  • Communicate the surprise

Example surprises:

CategoryExpectedUnexpected (Sticky)
Product launch"Introducing our new feature""We removed your favorite feature. Here's why."
Statistics"Obesity is growing""A bag of movie popcorn has more fat than a bacon-and-eggs breakfast, Big Mac and fries, and steak dinner — combined"
Value prop"Save money on insurance""15 minutes could save you 15%" (specific, unexpected)

Curiosity gaps:

  • Open a gap in knowledge → create desire to fill it
  • "Before I tell you the answer, let me ask..."
  • Mystery format: Present a puzzle, delay the resolution
  • Challenge assumptions: "You think X, but actually Y"

Creating curiosity gaps:

TechniqueHow It WorksExample
QuestionAsk what they don't know"What's the #1 reason startups fail?"
PredictionAsk them to predict"How many X do you think...?"
MysteryPresent a puzzle"Nordstrom once refunded a set of tires. They don't sell tires."
ChallengeViolate assumptions"Everything you know about X is wrong"

Anti-pattern: Gimmicky surprise without substance. The surprise must connect to the core message.

See: references/unexpected.md for pattern-breaking techniques.

3. Concrete

Core concept: Use sensory language and specific details instead of abstract concepts.

Abstract kills memorability. The more concrete and specific your idea, the stickier it becomes.

Abstract vs. Concrete:

AbstractConcrete
"Improve customer experience""Customers get their order in 30 minutes, still hot"
"Increase engagement""Users open the app 8 times a day"
"Optimize efficiency""Reduce report generation from 4 hours to 10 minutes"
"World-class support""Call us and a human answers in under 60 seconds"
"Scalable solution""Handle 10,000 users on day one without code changes"

The Velcro theory of memory:

  • Concrete ideas have more "hooks" for memory
  • "Bicycle" is easier to remember than "vehicle" (you can picture it)
  • Sensory details create mental images

Techniques for concreteness:

TechniqueHow It WorksExample
Specific numbersReplace "a lot" with exact figures"2,347 customers" not "thousands"
Sensory languageEngage senses"Crispy, not crunchy"
Concrete exampleReplace category with instance"Like John, a 35-year-old teacher in Denver"
DemonstrationShow, don't tellProduct demo > feature list
Before/afterTangible transformation"Before: 4 hours. After: 10 minutes."

Application to product messaging:

  • Features → Outcomes (what it does → what changes for user)
  • Percentages → Real numbers ("saves 40%" → "saves 16 hours/month")
  • Categories → Specific examples ("restaurants" → "pizza shops in Brooklyn")

See: references/concrete.md for concreteness exercises.

4. Credible

Core concept: Help people believe your idea using internal and external credibility.

External credibility:

SourceHow It WorksExample
AuthoritiesExpert endorsement"Recommended by Harvard Business Review"
Anti-authoritiesReal people with experience"Here's what a customer with the same problem found"
CredentialsVerifiable achievements"10 years experience, SOC 2 certified"

Internal credibility (more powerful):

TechniqueHow It WorksExample
Vivid detailsSpecificity implies truth"On Tuesday at 3pm, in the conference room on the 4th floor..."
StatisticsBut make them human-scaleNot "$1B market" but "1 in 4 businesses"
The Sinatra TestOne example so good it proves everything"If I can make it there, I can make it anywhere"
Testable credentialLet them verify"Try it free for 14 days"
Human-scale statisticsRelate numbers to experienceNot "10TB of data" but "every book ever written, 100 times"

The Sinatra Test:

  • One reference so impressive it handles all objections
  • "We secured the White House" = instant security credibility
  • "We handle Super Bowl traffic" = instant scalability credibility
  • "Used by Apple, Google, and Microsoft" = instant quality credibility

Making statistics sticky:

  • Don't: "37 grams of saturated fat"
  • Do: "More saturated fat than a Big Mac, fries, and milkshake combined"
  • Rule: Put statistics in a context people understand

See: references/credible.md for credibility-building techniques.

5. Emotional

Core concept: Make people feel something. People act on emotion, not analysis.

Mother Teresa principle: "If I look at the mass, I will never act. If I look at the one, I will."

Key insight: Statistics numb. Stories about individuals inspire action.

Emotional appeals:

ApproachHow It WorksExample
Individual focusOne person's story > statistics"Meet Sarah, who..." > "10,000 people affected"
Self-interest"What's in it for me?"WIIFM (features → personal benefits)
Identity"What would someone like me do?""Texans don't litter" (Don't Mess with Texas)
Maslow's hierarchyAppeal to the right levelSecurity, belonging, esteem, self-actualization

The identity approach:

  • People make decisions based on identity, not calculation
  • "What would a person like me do in this situation?"
  • Frame your product as consistent with who they want to be

Examples:

Identity FrameProductMessage
"I'm an innovative leader"SaaS tool"For teams that move fast"
"I care about my health"Food product"Made with ingredients you can pronounce"
"I'm a serious professional"B2B service"The tool Fortune 500 CTOs rely on"

Avoiding the "semantic stretch":

  • Don't over-abstract the emotion
  • "Support the troops" > "Support our national defense infrastructure"
  • Keep it personal and specific

See: references/emotional.md for emotional appeal frameworks.

6. Stories

Core concept: Stories are flight simulators for the brain. They teach people how to act.

Why stories work:

  • Simulate experience (mental rehearsal)
  • Inspire action (not just understanding)
  • Are memorable (narrative structure)
  • Bypass resistance (people don't argue with stories)

Three story plots that work:

PlotStructureWhen to UseExample
ChallengeProtagonist overcomes obstacleInspire courage, perseverance"We started in a garage..."
ConnectionPeople bridging a gapInspire tolerance, teamwork"A customer helped another customer..."
CreativityNovel solution to problemInspire innovation, thinking"We tried X, Y, Z... then discovered..."

Story structure for product messaging:

  1. Character: Who is the customer? (relatable)
  2. Problem: What challenge did they face? (emotional)
  3. Journey: What did they try? (concrete)
  4. Solution: How did your product help? (specific)
  5. Outcome: What changed? (measurable + emotional)

Example:

"Sarah ran a 10-person design agency. Her team spent 4 hours every Friday compiling client reports from 5 different tools. She'd tried hiring an intern, building spreadsheets, even a custom tool. Nothing worked. Then she found [Product]. Now reports generate in 10 minutes. Last Friday, her team left at 3pm for the first time in years."

Spotting stories in the wild:

  • Customer support tickets (problems + resolutions)
  • Sales calls (objections + breakthroughs)
  • User interviews (before/after moments)
  • Internal Slack (team wins)

See: references/stories.md for story templates and collection methods.

The Curse of Knowledge

The biggest enemy of sticky ideas.

Definition: Once you know something, you can't imagine not knowing it.

How it manifests:

  • Using jargon your audience doesn't know
  • Skipping context that seems "obvious"
  • Assuming your audience sees the same things you do
  • Over-abstracting because you know the specifics

Solutions:

  • Test messaging with outsiders (not your team)
  • Use concrete language, not abstractions
  • Tell stories, not bullet points
  • Ask: "Would my mom understand this?"

Sticky Messaging Audit

Rate your message on each principle:

PrincipleQuestionScore (1-10)
SimpleIs there ONE clear core message?
UnexpectedDoes it break a pattern or create curiosity?
ConcreteCan you picture it? Are there specific details?
CredibleWhy should someone believe this?
EmotionalDoes it make you feel something?
StoriesIs there a narrative or character?

Scoring:

  • 50-60: Extremely sticky (rare, aim for this)
  • 35-49: Strong (most good messaging lands here)
  • 20-34: Average (forgettable, needs work)
  • Below 20: Won't stick (fundamental rework needed)

Applying SUCCESs to Product

Landing Pages

  • Simple: One clear value proposition above the fold
  • Unexpected: Counterintuitive claim or statistic
  • Concrete: Specific outcome ("save 4 hours/week" not "save time")
  • Credible: Customer logos, specific testimonials
  • Emotional: Customer story or pain point
  • Stories: Customer transformation narrative

Product Demos

  • Simple: Show ONE core workflow, not every feature
  • Unexpected: Start with the "aha moment" not a tour
  • Concrete: Use real data, not "Lorem ipsum"
  • Credible: Show how [specific company] uses it
  • Emotional: Connect to the pain they feel today
  • Stories: "Let me show you what happens when [customer] has this problem..."

Onboarding

  • Simple: One action per screen
  • Unexpected: Delight with quick win early
  • Concrete: Show real results, not abstract promises
  • Credible: "Join 5,000 teams already using..."
  • Emotional: Celebrate first success
  • Stories: "Here's how [user] got started..."

Common Mistakes

MistakeWhy It FailsFix
Burying the leadCore message lost in detailsCommander's Intent: what's the ONE thing?
Too abstractNothing to rememberReplace every abstraction with a concrete example
Feature listingNo emotional connectionTell customer stories, show transformations
JargonCurse of KnowledgeTest with outsiders
Statistics without contextNumbers don't stickMake stats human-scale and relatable

Quick Diagnostic

Audit any message:

QuestionIf NoAction
Can I state the core in one sentence?Too complexFind Commander's Intent
Would this surprise someone?Predictable = forgettableFind the counterintuitive angle
Can I picture it happening?Too abstractAdd specific, sensory details
Why should someone believe this?No credibilityAdd proof, examples, Sinatra Test
Does it make me feel something?Purely logicalFocus on one person, not statistics
Is there a story?List of factsWrap in character + problem + resolution

Reference Files

Further Reading

This skill is based on Chip and Dan Heath's research on sticky ideas. For the complete framework:

About the Authors

Chip Heath is a professor at Stanford Graduate School of Business, and Dan Heath is a senior fellow at Duke University's CASE center. Together they have written four New York Times bestsellers. *Made to Stick* spent over 2 years on the bestseller list. Their research spans organizational behavior, decision-making, and how to make ideas have lasting impact. The SUCCESs framework is used by educators, marketers, nonprofits, and product teams worldwide.

平台分布

Codex

31.96%
按下载量换算3,424

Claude

30.76%
按下载量换算3,295

Cursor

19.54%
按下载量换算2,093

Gemini CLI

9.22%
按下载量换算988

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills