Token导航 LogoToken导航

Agent Skill · 研究检索

user-discovery-research(用户发现研究)

user-discovery-research 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。

user-discovery-research 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;许可证为 MIT,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 41,772。

总安装 1,140周安装 48GitHub Stars 2下载 399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
MIT

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:user-discovery-research(用户发现研究)
来源仓库:https://github.com/pluginagentmarketplace/custom-plugin-product-manager
仓库路径:skills/user-discovery-research
安装命令:
npx skills add https://github.com/pluginagentmarketplace/custom-plugin-product-manager --skill user-discovery-research
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/pluginagentmarketplace/custom-plugin-product-manager --skill user-discovery-research

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。

用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。适用于根据关键词或任务场景进行信息检索和筛选的场景。通过 npx skills add 命令从指定 GitHub 仓库安装使用。安装前需确认权限范围和维护状态,注意可能触发联网或文件操作。

user-discovery-research 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/user-discovery-research

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/pluginagentmarketplace/custom-plugin-product-m…;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

user-discovery-research 怎么安装?

在终端执行 npx skills add https://github.com/pluginagentmarketplace/custom-plugin-product-manager --s…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

user-discovery-research 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。

user-discovery-research 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。

user-discovery-research 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

user-discovery-research 用的人多吗?

公开数据是累计安装 1,140 次,近一周安装 48 次,下载 399 次,GitHub 2 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。

user-discovery-research 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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 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。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 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
user-discovery-research

用户发现研究 · 本页

1,1402只读MIT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

User Discovery & Research Skill

Become an expert at understanding customers deeply. Master interview techniques, extract meaningful insights, create accurate personas, and map customer journeys.

Interview Mastery

30-60 Minute Interview Structure

Opening (5 min) - Build rapport

  • "Thanks for time. I want to understand your experience with [topic]."
  • Warm conversation (not a interrogation)
  • Note-taking permission, recording permission

Context (10 min) - Background

  • "Tell me about your role..."
  • "What does a typical day look like?"
  • "What tools do you use?"
  • "Who do you work with?"

Deep Dive (15-20 min) - Core problem

  • "Walk me through how you currently [solve X]..."
  • "Tell me about the last time [problem] happened..."
  • "What frustrates you most about [current solution]?"
  • "How often does [issue] occur?"

Impact Questions (5-10 min) - Consequence

  • "What's the impact when [problem] happens?"
  • "How much time do you spend on [workaround]?"
  • "What's the cost if [issue] isn't resolved?"

Exploration (5-10 min) - Future state

  • "What would make your life easier?"
  • "Describe your ideal [solution]..."
  • "What would you be willing to pay to solve [problem]?"

Closing (3-5 min) - Wrap up

  • "Is there anything else important?"
  • "Can I follow up?"
  • "Can you introduce me to others?"

Interview Techniques

The Socratic Method

  • Ask questions instead of proposing solutions
  • Let them discover insights
  • Ask "Why?" 5 times to get root cause
  • Example:

- Q: "Why do you use spreadsheets?" - A: "It's what we have" - Q: "Why not use a database?" - A: "Don't have budget" - Q: "Why no budget?" - A: "ROI wasn't clear" - → Real problem: ROI uncertainty

Listen 70%, Talk 30%

  • You have two ears, one mouth
  • Let silence linger (they fill it)
  • Take notes but stay engaged
  • Nod and show genuine interest

Avoid Leading Questions

  • ❌ "Don't you think mobile is important?"
  • ✅ "What devices do you primarily use?"
  • ❌ "Wouldn't X solve your problem?"
  • ✅ "What would an ideal solution look like?"

Look for Emotions

  • Frustration, excitement, anxiety
  • These reveal true pain points
  • "I can see this is frustrating. Tell me more..."

Interview Logistics

Recruitment:

  • Target: 20-25 interviews minimum
  • Mix: power users, light users, non-users
  • Incentive: $50-100 or product credit
  • Screening: Ensure they fit target segment

Recording & Notes:

  • Always record (with permission)
  • Take live notes (rough OK)
  • Timestamps for quotes
  • Video is better than audio if possible

Debrief Immediately:

  • Write key insights within 24 hours
  • While memory fresh, fill in gaps
  • Highlight best quotes
  • Note any bias you notice

Synthesis & Analysis

Affinity Mapping

Process:

  1. Write every quote/insight on sticky note
  2. Group similar insights
  3. Find themes
  4. Rank by frequency/impact
  5. Extract top 10-15 themes

Example themes:

  • Time pressure (everyone mentions)
  • Tool switching costs (8 of 20 people)
  • Lack of automation (7 of 20 people)
  • Data fragmentation (15 of 20 people)

Persona Development

5 Element Persona:

  1. Demographics

- Role, company size, industry - Education, experience level - Geographic location - Company revenue/growth stage

  1. Goals & Motivations

- What's their job-to-be-done? - How is success measured? - What drives them? - What keeps them up at night?

  1. Challenges & Pain Points

- Top 3 problems (research-backed) - Current workarounds - Cost of problem (time, money, stress) - What they've tried

  1. Current Solutions

- What tools do they use? - How do they work together? - What's missing? - Tech stack

  1. Quote & Story

- Real, powerful quote - Day-in-the-life scenario - Decision journey story

Persona Document Format:

  • 1-2 pages max
  • Include photo (stock image)
  • Real quotes (3-5)
  • Scenario/story
  • Key behaviors
  • Success metrics

Journey Mapping

Current State Journey (As-Is)

Stages:

  1. Awareness - Realizes problem exists
  2. Consideration - Researches solutions
  3. Purchase - Buys solution
  4. Onboarding - Learns to use
  5. Usage - Day-to-day work
  6. Support - Gets help
  7. Renewal - Continues or leaves

For Each Stage Map:

Actions - What customer does

  • Searches online
  • Talks to colleagues
  • Evaluates options
  • etc.

Emotions - How they feel

  • Frustrated (searching)
  • Cautious (evaluating)
  • Confident (purchased)
  • Overwhelmed (onboarding)
  • etc.

Touchpoints - Interactions

  • Website
  • Sales calls
  • Email
  • Product UI
  • Support chat
  • etc.

Pain Points - Frustrations

  • Hard to find info
  • Slow response times
  • Confusing UI
  • etc.

Opportunities - Where you can help

  • Better documentation
  • Faster support
  • Simpler onboarding
  • etc.

Research Methods

Quantitative

Surveys

  • Reach: 100+ people
  • Speed: 2-3 days
  • Depth: Medium
  • Best for: Validation, scale

Net Promoter Score (NPS)

  • "How likely to recommend?" (0-10)
  • Followers: 9-10
  • Passives: 7-8
  • Detractors: 0-6
  • Formula: %Followers - %Detractors

Qualitative

Interviews

  • Reach: 15-25 people
  • Speed: 2-3 weeks
  • Depth: Very high
  • Best for: Understanding "why"

Focus Groups

  • Reach: 8-12 people
  • Speed: 1-2 weeks
  • Depth: High
  • Bonus: Group dynamic reveals thinking

Observations

  • Reach: 3-5 people
  • Speed: 2-3 days
  • Depth: Very high
  • Best for: Seeing vs. hearing

Mixed Methods

Combine research methods:

  • Start with interviews (understand why)
  • Then survey (validate with 100+)
  • Iterate based on findings

Key Insights Extraction

Five Key Metrics

  1. Problem Severity

- How badly does it hurt? - Frequency of problem - Impact magnitude - Urgency to solve

  1. Current Solution Satisfaction

- Is current approach acceptable? - What's missing? - Workarounds needed? - Would switch if better option?

  1. Willingness to Pay

- What would they pay for solution? - Price sensitivity? - ROI expectations?

  1. Decision Making

- Who decides? - Who influences? - Budget process? - Timeline?

  1. Switching Costs

- What prevents switching? - Integration challenges? - Learning curve acceptable? - Migration effort?

Top 5 Questions to Answer

From your research:

  1. What's the #1 problem?
  2. Who has the problem most acutely?
  3. How do they solve it today?
  4. What would ideal solution look like?
  5. Would they pay for it, and how much?

Troubleshooting

Yaygın Hatalar & Çözümler

HataOlası SebepÇözüm
Düşük interview katılımıYanlış incentive$50-100 gift card
Yüzeysel insightsLeading questions"Why?" 5x sor
Conflicting feedbackMixed segmentsSegment-based analysis
Persona genericInsufficient data+10 interview

Debug Checklist

[ ] Sample size yeterli mi? (min 15-20)
[ ] Questions leading değil mi?
[ ] Recording consent alındı mı?
[ ] Synthesis 24h içinde yapıldı mı?
[ ] Bias check yapıldı mı?

Recovery Procedures

  1. Low Participation → Adjust incentive or channel
  2. Conflicting Data → Segment by user type
  3. Interviewer Bias → Add second interviewer

Master deep customer understanding and let it guide your product decisions!

平台分布

Claude Code

28.7%
按下载量换算115

Antigravity

23.07%
按下载量换算92

OpenCode

19.1%
按下载量换算76

Codex

12.97%
按下载量换算52

Gemini CLI

8.41%
按下载量换算34

windsurf

3.39%
按下载量换算14

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。

继续浏览同类 Skills