chuinb
This skill transforms users into confident industry insiders through immersive, research-driven learning experiences. It should be used when users want to quickly master an unfamiliar industry, field, skill, or domain — whether for professional networking, investment decisions, career transitions, or pure curiosity. The skill combines Feynman Technique, First Principles thinking, and 80/20 analysis with deep web research to generate beautifully formatted, media-rich learning notes with interactive flashcards and quizzes. Triggers include: "help me understand [industry]", "I need to learn about [field] quickly", "make me an insider in [domain]", "行业速成", "快速掌握", "深度学习笔记", "/chuinb", "/master", or any request to rapidly acquire domain expertise.
- Installs
- 63
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/yizhiyanhua-ai/chuinb-skill --skill chuinb- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads chuinb automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Industry Mastery: 行业速成大师
Transform from outsider to insider in hours, not months.
This skill creates immersive, research-backed learning experiences that make users feel like industry veterans. It combines proven learning methodologies with real-time web research to deliver knowledge that sticks.
Core Philosophy
The Three Pillars
┌─────────────────────────────────────────────────────────────────┐
│ INDUSTRY MASTERY │
├─────────────────────────────────────────────────────────────────┤
│ │
│ 🧠 FEYNMAN TECHNIQUE ⚛️ FIRST PRINCIPLES │
│ "If you can't explain "Boil everything down │
│ it simply, you don't to fundamental truths, │
│ understand it well enough" then reason up from there" │
│ │
│ 📊 80/20 PARETO │
│ "20% of knowledge delivers │
│ 80% of practical value" │
│ │
└─────────────────────────────────────────────────────────────────┘
⚠️ CRITICAL: Execution Flow (MUST FOLLOW)
Overview
┌─────────────────────────────────────────────────────────────────┐
│ EXECUTION FLOW │
├─────────────────────────────────────────────────────────────────┤
│ │
│ Phase 1: User Profiling ────────────────────────────────── │
│ Ask 3 questions about goal, background, time │
│ ↓ │
│ Phase 2: Deep Research ─────────────────────────────────── │
│ WebSearch + WebFetch for content │
│ ↓ │
│ Phase 3: Media Acquisition (MANDATORY) ─────────────────── │
│ Download images + videos + generate AI images │
│ ↓ │
│ Phase 4: Ask Save Path ─────────────────────────────────── │
│ Use AskUserQuestion to get save location │
│ ↓ │
│ Phase 5: Generate & Save ───────────────────────────────── │
│ Create markdown file with embedded media │
│ ↓ │
│ Phase 6: Interactive Follow-up ─────────────────────────── │
│ Offer deep dives, practice scenarios │
│ │
└─────────────────────────────────────────────────────────────────┘
Phase 1: User Profiling (MANDATORY FIRST STEP)
Before ANY research begins, gather user context through conversational questions:
┌─────────────────────────────────────────────────────────────────┐
│ 🎯 USER PROFILING QUESTIONS │
├─────────────────────────────────────────────────────────────────┤
│ │
│ 1. 「你的目标」What do you want to achieve? │
│ □ 社交谈资 (Casual conversation) │
│ □ 职业转型 (Career transition) │
│ □ 投资决策 (Investment decisions) │
│ □ 合作洽谈 (Business collaboration) │
│ □ 纯粹好奇 (Pure curiosity) │
│ │
│ 2. 「当前背景」What's your current profession/background? │
│ (This helps tailor analogies and explanations) │
│ │
│ 3. 「时间预算」How much time can you invest? │
│ □ 30分钟速览 (Quick overview) │
│ □ 2小时深入 (Deep dive) │
│ □ 持续学习 (Ongoing learning) │
│ │
└─────────────────────────────────────────────────────────────────┘
Present these questions conversationally, not as a form. Adapt based on context clues in the user's initial request.
Phase 2: Deep Research
Execute comprehensive web research covering:
-
Industry Fundamentals
- Core business models and value chains
- Key players (companies, organizations)
- Market size and growth trends
-
Key Figures & Events
- Influential people (founders, thought leaders, critics)
- Historical milestones and turning points
- Recent news and developments
-
Professional Vocabulary
- Industry jargon and acronyms
- Insider phrases and expressions
- Common misconceptions to avoid
-
Real Cases & Stories
- Success stories with specific details
- Notable failures and lessons learned
- Current controversies or debates
Research Tools:
- Use
WebSearchfor current information, news, trends - Use
WebFetchfor detailed article content
2.1 Source Credibility Guardrails (MANDATORY)
在整理研究结果前，必须执行以下可信度约束：
- 最少来源数量：至少引用 5 个有效来源
- 来源类型配比（至少覆盖 2 类）：
- 官方/监管/机构报告（公司财报、政府、协会、咨询机构）
- 主流媒体或行业媒体
- 专业研究文章或高质量数据库
- 时间窗口要求：
- 至少 2 条信息来自最近 12 个月
- 涉及“最新进展”时，优先最近 90 天来源
- 可追溯性：关键结论后面必须附来源链接或出处说明
2.2 Conflict Handling Protocol
当不同来源出现冲突信息时，遵循以下流程：
- 明确列出冲突点（例如市场规模、增长率、关键事件日期）
- 给出每个版本的来源与时间
- 解释可能的差异原因（统计口径、地区范围、时间维度）
- 给出“当前最可信结论 + 不确定性说明”
2.3 Citation & Evidence Rules (MANDATORY)
- 严禁编造来源：找不到可信出处时，明确写“暂无权威来源支持”
- 关键结论必须可追溯：核心观点后附
[来源名, 日期] + 链接 - 数据必须带时间戳：市场规模、增速、排名等信息需注明统计年份
推荐在输出中使用如下格式：
### 关键结论与证据
1. 结论 A（置信度：高）
- 证据 1: [来源名称, 2026-03-10](https://example.com)
- 证据 2: [来源名称, 2025-11-22](https://example.com)
- 备注: 不同机构口径略有差异，已按“全球口径”统一
2.4 Reference Module Routing
按主题选择最小必要参考文件，避免过载：
- 术语与表达：
references/jargon-patterns.md - 行业分类研究框架：
references/industry-templates.md - 学习方法解释：
references/learning-theories.md
仅加载与当前用户目标相关的部分，不必每次全量展开。
Phase 3: Media Acquisition (MANDATORY - DO NOT SKIP)
⚠️ THIS PHASE IS REQUIRED FOR EVERY EXECUTION
媒体素材是让学习笔记生动有力的关键。必须执行媒体获取流程，但数量和类型根据内容需要灵活调整。
3.1 媒体获取原则
核心原则：根据内容类型选择最合适的获取方式
┌─────────────────────────────────────────────────────────────────┐
│ 媒体获取决策树 │
├─────────────────────────────────────────────────────────────────┤
│ │
│ 需要什么类型的图片？ │
│ │ │
│ ├─→ 事实性图片 ──→ 优先网络下载真实图片 │
│ │ • 人物照片（创始人、专家、名人） │
│ │ • 产品图片、公司 Logo │
│ │ • 剧照、海报、新闻配图 │
│ │ • 历史事件照片 │
│ │ • 实物图片（咖啡豆、芯片、汽车等） │
│ │ │
│ └─→ 概念性图片 ──→ 使用 AI 生成 │
│ • 价值链/生态系统图 │
│ • 流程图、关系图 │
│ • 抽象概念的视觉化 │
│ • 数据可视化、信息图 │
│ • 氛围图、风格示意图 │
│ │
└─────────────────────────────────────────────────────────────────┘
3.2 媒体数量指南（灵活调整）
不设硬性数量限制，根据内容丰富度和主题特点决定：
| 内容类型 | 建议媒体配置 | 说明 |
|---|---|---|
| 人物密集型（如影评、商业领袖） | 3-5 张人物照片 + 1-2 个访谈视频 | 优先下载真实照片 |
| 概念密集型（如金融、技术） | 2-4 张概念图 + 1-2 个解释视频 | 优先 AI 生成 |
| 案例密集型（如商业分析） | 2-3 张案例图 + 1-2 张流程图 | 混合使用 |
| 视觉艺术类（如电影、设计） | 4-6 张剧照/作品图 + 1-2 个片段 | 优先下载真实图片 |
3.3 图片获取策略
策略 A：事实性图片 → 优先网络下载
适用场景：
- 人物照片（创始人、CEO、专家、艺术家、导演、演员等）
- 产品图片、公司 Logo、品牌视觉
- 电影剧照、海报、专辑封面
- 新闻事件配图、历史照片
- 实物图片（食物、设备、建筑等）
获取方式：
# 方式 1: 使用 media-downloader（需要 API Key）
python ~/.claude/skills/media-downloader/media_cli.py image "关键词" -n 数量 -o 输出目录
# 方式 2: 通过 WebSearch 找到图片 URL，然后下载
# 搜索关键词示例：
# "[人名] portrait photo"
# "[公司名] logo high resolution"
# "[电影名] movie poster"
# "[产品名] product image"
搜索关键词模板：
人物照片: "[姓名] portrait" / "[姓名] headshot" / "[姓名] photo"
公司Logo: "[公司名] logo png" / "[公司名] brand"
电影海报: "[电影名] movie poster" / "[电影名] official poster"
剧照: "[电影名] still" / "[电影名] scene" / "[电影名] screenshot"
产品图: "[产品名] product photo" / "[产品名] official image"
策略 B：概念性图片 → 使用 AI 生成
适用场景：
- 行业价值链、生态系统图
- 业务流程图、工作流程
- 抽象概念的视觉化表达
- 关系图、层级图
- 氛围图、风格示意图
- 找不到合适真实图片时的备选
使用 zimage-skill 生成：
MODELSCOPE_API_KEY="your-key" python3 ~/.claude/skills/zimage-skill/generate.py "prompt" "output.jpg"
Prompt 模板：
# 流程图/价值链
"[主题] value chain diagram, minimalist infographic style, [color] color scheme,
professional business design, clean flat design, white background"
# 概念图
"[概念] concept visualization, modern illustration style, simple and clear,
professional corporate design"
# 氛围图
"[主题] aesthetic, cinematic atmosphere, [风格描述], professional photography style"
# 生态系统图
"[行业] ecosystem diagram, showing key players and relationships,
minimalist business infographic, clean design"
策略 C：备选方案
当以上方式都失败时：
- 提供外部链接：
[查看图片](url) - 使用文字描述代替
- 使用 Mermaid 图表（适用于流程图）
3.4 视频获取策略
使用 media-downloader 下载 YouTube 视频：
python ~/.claude/skills/media-downloader/media_cli.py youtube "URL" -o "输出目录" --end 120
视频类型与搜索关键词：
| 视频类型 | 搜索关键词模板 | 适用场景 |
|---|---|---|
| 解释性视频 | "[概念] explained", "how [X] works" | 复杂概念讲解 |
| 入门视频 | "[行业] 101", "[行业] for beginners" | 行业入门 |
| 人物访谈 | "[人名] interview", "[人名] talk" | 人物介绍 |
| TED 演讲 | "[主题] TED talk" | 思想启发 |
| 纪录片片段 | "[主题] documentary" | 深度内容 |
| 新闻报道 | "[事件] news report" | 时事案例 |
视频要求：
- 时长：根据内容价值灵活裁剪（建议 60-180 秒）
- 内容：与主题直接相关
- 质量：至少 720p
3.5 媒体文件命名规范
人物照片: person-[姓名拼音或英文].jpg
概念图: diagram-[描述].jpg
剧照/海报: poster-[作品名].jpg / still-[作品名].jpg
案例图: case-[案例名].jpg
产品图: product-[产品名].jpg
视频: video-[主题].mp4
3.6 媒体嵌入格式 (Obsidian)
图片: ![[media/filename.jpg]]
视频: ![[media/filename.mp4]]
带说明: ![[media/filename.jpg|这是图片说明]]
3.7 媒体获取检查清单
在完成媒体获取后，确认：
- 所有提到的关键人物都有对应图片（真实照片优先）
- 核心概念有视觉化表达（AI 生成或真实图片）
- 至少有 1 个相关视频片段
- 图片和视频数量与内容丰富度匹配
- 所有媒体文件已下载到本地 media 文件夹
- 文件命名清晰规范
3.8 媒体失败降级策略（MANDATORY）
当媒体获取失败时，不允许直接跳过，必须按顺序降级：
- 事实图失败：尝试 AI 生成替代图，并明确标注“AI 生成示意图”
- AI 图失败：改用 Mermaid/ASCII 图或文字结构化描述
- 视频失败：提供高质量外链 + 60-120 秒关键内容摘要
3.9 版权与合规提示（MANDATORY）
在文档的“延伸阅读”或末尾补充：
- 媒体来源说明（来源平台/作者/链接）
- 版权状态提示（仅学习参考或可商用）
- 商业使用风险提醒（建议优先使用可复用许可素材）
Phase 4: Ask Save Path (MANDATORY)
⚠️ 在生成内容之前，必须询问用户保存路径
使用 AskUserQuestion 工具询问用户：
问题: "请告诉我你想把学习笔记保存到哪里？"
选项:
1. 当前目录 (Current directory)
2. 桌面 (Desktop)
3. 自定义路径 (Custom path)
如果用户选择自定义路径:
- 等待用户输入完整路径
- 验证路径是否存在，不存在则创建
默认文件结构:
[用户指定路径]/
├── [主题]速成指南.md # 主文档
└── media/ # 媒体文件夹
├── diagram-*.jpg
├── person-*.jpg
├── case-*.jpg
└── video-*.mp4
Phase 5: Content Generation & Save
5.1 Output Structure Template
# [Industry/Field Name] 行业速成指南
> 🎯 **你的学习目标**: [Personalized based on user profile]
> ⏱️ **预计阅读时间**: X 分钟
> 📅 **生成日期**: YYYY-MM-DD
---
## 一句话看懂这个行业
[Feynman-style explanation in ONE sentence that a 12-year-old could understand]
---
## 第一性原理：行业的本质
[Break down to fundamental truths. What problem does this industry solve? Why does it exist?]
![[media/diagram-value-chain.jpg]]
*行业价值链图解*
### 核心价值链
[Visual diagram or clear explanation of how value flows]
### 关键驱动因素
[What makes this industry tick? 3-5 key factors]
---
## 行话速成：像内行人一样说话
| 术语 | 含义 | 使用场景 |
|------|------|----------|
| Term 1 | Meaning | When to use |
| Term 2 | Meaning | When to use |
| ... | ... | ... |
### 常用表达
- "[Insider phrase 1]" — 意思是...
- "[Insider phrase 2]" — 用于...
### 新手常犯的错误
- ❌ 不要说"..." → ✅ 应该说"..."
---
## 必知人物
### [Name 1] — [Title/Role]
> "[Famous quote]"
[Brief bio and why they matter]
### [Name 2] — [Title/Role]
...
---
## 经典案例
### 案例一：[Success/Failure Story Title]
**背景**: ...
**过程**: ...
**结果**: ...
**启示**: [Key takeaway in user's professional context]
![[media/case-example.jpg]]
---
## 精选视频片段
### [Video Title]
![[media/video-explanation.mp4]]
> 📌 **关键点**: [1-2 sentence summary of why this matters]
---
## 闪念卡片 (Flashcards)
<details>
<summary>🔮 点击展开卡片 1</summary>
**Q: [Question]**
---
**A: [Answer]**
</details>
[Generate 5-10 flashcards covering key concepts]
---
## 自测问答
### 问题 1
[Scenario-based question]
<details>
<summary>💡 查看答案</summary>
[Answer with explanation]
</details>
[Generate 3-5 quiz questions]
---
## 行动清单
基于你的目标「[user goal]」，建议的下一步：
- [ ] [Actionable item 1]
- [ ] [Actionable item 2]
- [ ] [Actionable item 3]
---
## 延伸阅读
- [Resource 1](url) — 推荐理由
- [Resource 2](url) — 推荐理由
---
> 💡 **学习小贴士**: [Personalized tip based on user's background]
5.2 Save Files
- 创建 media 文件夹
- 将所有媒体文件保存到 media 文件夹
- 保存主 markdown 文件
5.3 Scenario Modes (Output Adaptation)
根据用户目标自动选择输出重点（可显式声明 mode）：
mode=networking：强调谈资、人物、热点话题mode=interview：强调高频问答、结构化表达、案例拆解mode=investment：强调商业模式、增长驱动、风险与竞争格局mode=learning：强调概念体系、学习路径、记忆卡片
5.4 Output Quality Rubric (Self-check)
交付前按 1-5 分自评并简要补强薄弱项：
- 清晰度（小白可理解）
- 准确性（来源可信、可追溯）
- 实用性（可用于会议/面试/投资交流）
- 记忆性（闪卡与测验质量）
- 可读性（结构、排版、图文对应）
若总分 < 18，需至少补强 2 个低分维度后再交付。
Phase 6: Interactive Elements
After delivering the main note, offer interactive follow-ups:
-
Deep Dive Options
- "想深入了解 [specific topic] 吗？"
- "需要更多 [cases/videos/terminology] 吗？"
-
Practice Scenarios
- "假设你在 [场景]，对方问你 [问题]，你会怎么回答？"
- Provide feedback on user's responses
-
Knowledge Checks
- Generate additional flashcards on demand
- Create scenario-based quizzes
-
Connection Building
- "这和你的 [user's profession] 有什么关联？"
- Help user find bridges between new knowledge and existing expertise
Content Guidelines
Writing Style
- Conversational yet professional — Like a knowledgeable friend explaining things
- Rich in analogies — Connect new concepts to familiar ones (based on user's background)
- Concrete over abstract — Always include specific examples, numbers, names
- Visual thinking — Use diagrams, tables, and formatting liberally
Personalization Markers
Throughout the note, include personalized elements:
「结合你的[background]来理解」— Bridge to user's expertise「对于[goal]来说，关键是...」— Goal-oriented framing「这就像你熟悉的[analogy from user's field]」— Familiar analogies
Quality Checklist
Before delivering the note, verify:
- User profile questions were asked and answers incorporated
- At least 5 web searches performed for current information
- At least 5 valid references collected and traceable
- At least 2 references are from the last 12 months
- At least 2 source types covered (official/media/research)
- Conflicting facts (if any) are explicitly reconciled
- Minimum 3 real cases with specific details
- 10+ industry terms explained
- 2+ key figures profiled
- 5+ flashcards generated
- 3+ quiz questions created
- 媒体获取完成：
- 事实性图片（人物、剧照等）优先从网络下载真实图片
- 概念性图片（流程图、价值链）使用 AI 生成
- 图片数量与内容丰富度匹配（不设硬性限制）
- 至少 1 个相关视频片段
- User was asked for save path before saving
- Markdown file saved with embedded media
- Media source and copyright/compliance note included
- Output mode selected or inferred from user goal
- Rubric self-check completed (target total >= 18)
- Personalization markers present throughout
- Actionable next steps provided
Tool Integration
Required Tools
| Tool | Purpose | When to Use |
|---|---|---|
WebSearch | 搜索行业信息 | 每次必用 |
WebFetch | 获取网页详细内容 | 每次必用 |
zimage-skill | 生成概念图 | 概念性图片（流程图、价值链等） |
media-downloader | 下载图片和视频 | 事实性图片 + YouTube 视频 |
AskUserQuestion | 询问保存路径 | 保存前必用 |
zimage-skill (AI 图片生成)
功能: 使用 AI 生成概念图、流程图等
使用方式:
MODELSCOPE_API_KEY="your-key" python3 ~/.claude/skills/zimage-skill/generate.py "prompt" "output.jpg"
Prompt 最佳实践:
"[主题] diagram/infographic, minimalist style, [color] color scheme,
professional design, clean flat design, white background"
media-downloader (媒体下载器)
功能: 下载 YouTube 视频并裁剪
使用方式:
# 下载并裁剪视频
python ~/.claude/skills/media-downloader/media_cli.py youtube "URL" -o "目录" --end 120
# 检查配置状态
python ~/.claude/skills/media-downloader/media_cli.py status
Example Triggers
- "帮我快速了解私募股权行业"
- "I need to understand the semiconductor industry by next week"
- "想成为咖啡行业的内行人"
- "Help me master the basics of venture capital"
- "下周要和动漫行业的人聊天，帮我速成"
- "/chuinb blockchain"
- "/master contemporary art"
Troubleshooting
zimage-skill 报错 "API Key required"
需要设置环境变量：
export MODELSCOPE_API_KEY="your-api-key"
获取 API Key: https://modelscope.cn/my/myaccesstoken
media-downloader 图片下载失败
需要配置图库 API Key：
export PEXELS_API_KEY="your-key"
export PIXABAY_API_KEY="your-key"
备选方案: 使用 zimage-skill 生成图片代替下载
YouTube 视频下载失败
确保已安装 yt-dlp：
pip install yt-dlp
FAQ
What does the chuinb skill do?
This skill transforms users into confident industry insiders through immersive, research-driven learning experiences. It should be used when users want to quickly master an unfamiliar industry, field, skill, or domain — whether for professional networking, investment decisions, career transitions, or pure curiosity. The skill combines Feynman Technique, First Principles thinking, and 80/20 analysis with deep web research to generate beautifully formatted, media-rich learning notes with interactive flashcards and quizzes. Triggers include: "help me understand [industry]", "I need to learn about [field] quickly", "make me an insider in [domain]", "行业速成", "快速掌握", "深度学习笔记", "/chuinb", "/master", or any request to rapidly acquire domain expertise.
How do I install chuinb?
Run npx skills add https://github.com/yizhiyanhua-ai/chuinb-skill --skill chuinb in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/chuinb/ (personal) or .claude/skills/chuinb/ (project).
Which AI agents can use chuinb?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is chuinb safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains chuinb?
It is published by yizhiyanhua-ai in yizhiyanhua-ai/chuinb-skill. It has 63 recorded installs.
Popular education & science skills
teach
mattpocock/skills
Teach the user a new skill or concept, within this workspace.
scaffold-exercises
mattpocock/skills
Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use when user wants to scaffold exercises, create exercise stubs, or set up a new course section.
paper-context-resolver
lllllllama/rigorpilot-skills
Rigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical gap and the task is to resolve a specific paper detail such as dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption from primary paper sources while recording conflicts. Do not use for general paper summary, repo scanning, environment setup, command execution, title-only paper lookup, or replacing README guidance by default.
caveman-learn
juliusbrussee/caveman
Act on a Caveman learn report - review the ranked token sinks, apply cost-lowering fixes with per-edit consent, and report what those fixes returned. Use when asked to lower an agent's token cost, what caveman has saved, to trim a heavy CLAUDE.md, or to offload re-pasted context into cavemem.
gsap-plugins
greensock/gsap-skills
Official GSAP skill for GSAP plugins — registration, ScrollToPlugin, ScrollSmoother, Flip, Draggable, Inertia, Observer, SplitText, ScrambleText, SVG and physics plugins, CustomEase, EasePack, CustomWiggle, CustomBounce, GSDevTools. Use when the user asks about a GSAP plugin, scroll-to, flip animations, draggable, SVG drawing, or plugin registration.
teach-impeccable
pbakaus/impeccable
DEPRECATED: Use /impeccable teach instead. This command has been folded into the impeccable skill.