Token导航 LogoToken导航

Agent Skill · 研究检索

youtube-producer(YouTube 制作人)

youtube-producer 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。

youtube-producer 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,权限为执行命令;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 59,803。

总安装 376周安装 16GitHub Stars 3下载 132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:youtube-producer(YouTube 制作人)
来源仓库:https://github.com/kkoppenhaver/cc-skills
仓库路径:skills/youtube-producer
安装命令:
npx skills add https://github.com/kkoppenhaver/cc-skills --skill youtube-producer
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/kkoppenhaver/cc-skills --skill youtube-producer

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。

支持 YouTube 内容生产流程的信息检索。适合创作者在研究选题或竞品分析时使用。通过 npx skills add 命令安装。需核实关键词覆盖范围与结果准确性。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

youtube-producer 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/youtube-producer

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/kkoppenhaver/cc-skills --skill youtube-produce…;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

    它需要执行命令,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

youtube-producer 怎么安装?

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

youtube-producer 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。

youtube-producer 需要什么权限?

权限级别是执行命令。安装流程涉及命令执行,可能通过 npx skills add https://github.com/kkoppenhaver/cc-skills --skill youtube-producer 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

youtube-producer 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。

youtube-producer 用的人多吗?

公开数据是累计安装 376 次,近一周安装 16 次,下载 132 次,GitHub 3 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。

youtube-producer 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。

先想清楚的地方

  • 需要执行命令,装到常用环境前要先评估影响面。
  • 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。

不太适合

  • 不能接受 Skill 在你机器上执行命令的人。
  • 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
youtube-producer

YouTube 制作人 · 本页

3763执行命令未明确通过

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

YouTube Producer

Transform vague video ideas into compelling, viral-ready YouTube content using proven frameworks from top technical creators Mark Rober and Aaron Francis.

Core Production Philosophy

The Two Essential Frameworks

Mark Rober's "Hiding the Vegetables": Make education so entertaining that viewers don't realize they're learning. Lead with spectacle, education follows naturally.

Aaron Francis's "Nugget + Packaging": Every video has an educational nugget (the technical concept) wrapped in compelling packaging (the relatable problem/story) that makes it clickable without being misleading.

Core Workflow

When a user provides a video idea, follow this production pipeline:

1. Identify the Nugget and Find the Packaging

First, extract the educational nugget - What technical concept or skill will viewers learn?

Then, find the compelling packaging by asking:

  • What problem does this solve?
  • What frustration does this address?
  • What curiosity does this satisfy?
  • What justice does this deliver?

Example transformations:

Educational NuggetCompelling Packaging
Virtual columns in MySQL"Stop freeloaders exploiting your SaaS"
Database indexing"Why your app is slow (and how to fix it)"
AI transformer architecture"How ChatGPT learned to lie"

2. Title Generation Using Proven Formulas

Apply these high-performing patterns:

Contrarian Commands (Aaron Francis):

  • "Stop using [common practice]"
  • "You don't need [assumed requirement]"
  • "[Common belief] is a lie"
  • Add parenthetical qualifiers for credibility: "(anymore)", "(for shipping)", "(or sooner)"

Superlative Hooks (Mark Rober):

  • "World's [Largest/Smallest/First] [Thing]"
  • Promises something viewers have never seen before

Question Format:

  • "Can [AI/You] Really [Seemingly Impossible Thing]?"
  • "What Happens When [Unexpected Scenario]?"
  • Poses questions viewers didn't know they had

Justice/Problem-Solving:

  • "I Used [Technology] to [Solve Annoying Problem]"
  • "How [Technology] Caught [Bad Actor]"
  • Engineering as superpower for satisfaction

Permission-Giving:

  • "You Don't Always Need [Complex Thing]"
  • Liberates viewers from dreaded complexity

3. The 30-Second Hook Structure

Combine both creators' approaches:

0-3 seconds (Visual Validation):

  • Immediately show visual proof of title's promise
  • Most impressive moment or result
  • No introductions, straight to spectacle

3-10 seconds (Establish Stakes/Problem):

  • Surface problem viewers didn't know they had
  • Create "curiosity loop" that needs closing
  • Frame uncertainty: "We're not sure if this will work..."

10-20 seconds (Identity & Credibility):

  • "If you're a [target audience]..."
  • Brief credibility marker
  • Why they should trust you

20-30 seconds (Promise & Roadmap):

  • Clear promise of what they'll gain
  • Preview the journey (not the answer)
  • Pattern interrupt: "But first..." or "However..."

4. Video Structure (The Three-Act Journey)

Act 1: Hook & Personal Stakes (0-2 minutes)

  • Open with spectacle or problem
  • Establish personal connection/frustration
  • Character introduction (name your subjects/tools)
  • Set expectations without revealing outcome

Act 2: The Journey (60-70% of video)

  • Document process including failures (Super Mario Effect)
  • Chunk into 2-3 minute segments with visual variety
  • Alternate high-energy reveals with educational explanations
  • Show iterations and problem-solving
  • Weave education naturally into narrative
  • Include "breathing room" after complex concepts

Act 3: Climactic Payoff (Final 20%)

  • Deliver the big reveal
  • Perspective shift (frustration to admiration)
  • Broader implications/lessons
  • Circle back to opening question
  • Single, clear call-to-action

5. YouTube Shorts Extraction

For each main video, identify 3-5 short moments:

The Hook Short (15-30s):

  • Most shocking moment without context
  • Title: Question the main video answers

The Correction Short (30-45s):

  • "Stop doing [common practice]"
  • Quick explanation of why it's wrong
  • Tease full solution in main video

The Permission Short (30-45s):

  • "You don't need [complex thing]"
  • Show simpler alternative
  • Link to full explanation

The Justice/Result Short (15-30s):

  • Show the satisfying payoff
  • Package thief getting glittered
  • AI catching the scammer

Thumbnail Design Principles

Mark Rober's 3-Element Rule

  • Subject (the build/experiment) - dominates frame
  • Creator (relatively small) - 57% happy/confident expression
  • Context/scale reference - shows magnitude

Aaron Francis's Clean Technical Style

  • Clean code/terminal screenshots
  • Minimal text overlay
  • Professional without desperation
  • Tool logos when relevant

Title-Thumbnail Synergy

  • Thumbnail poses visual riddle
  • Title completes the thought
  • Together create "intensified curiosity"
  • Never duplicate information

Production Philosophy

Quality Over Quantity

  • Mark Rober: ~10 videos/year, some taking 3+ years
  • Aaron Francis: Only content you're genuinely excited about
  • Focus: "For something to be remarkable, it has to be able to be remarked about"

The "Try Hard" Differentiator

  • High production value signals content value
  • The "1000 little things" compound into quality perception
  • Clean setup, hidden desktop icons, good lighting
  • Be yourself, but amplified 2 notches for camera energy

Character-Driven Technical Content

  • Name your AI models like Mark names squirrels
  • Document failures as part of the journey
  • Make viewers co-explorers, not students
  • Show personality and vulnerability

Retention Optimization Techniques

Visual Variety (The Chunk Method)

  • Switch between talking head and screen every 2-3 minutes
  • Creates natural engagement checkpoints
  • Prevents visual fatigue
  • Maintains "dopamine pumping"

Open Loops & Pattern Interrupts

  • Promise multiple payoffs throughout
  • "But there's a problem..." moments
  • Tease coolest part early
  • Use "However," "Yet," "But" for transitions
  • Never reveal answer immediately

The Super Mario Effect

  • Focus on the goal (princess), not failures (pits)
  • Show failures as learning, not embarrassment
  • Make content feel like playing, not testing
  • Results in 2.5x more persistence, 68% higher success

Creator Framework Selection

Use specific creator approaches based on content type:

Mark Rober Style for:

  • Large builds and experiments
  • Visual spectacle content
  • Family-friendly education
  • Justice/satisfaction narratives
  • See references/mark-rober-framework.md

Aaron Francis Style for:

  • Developer/technical education
  • Database and backend topics
  • Controversial technical opinions
  • Business problem solving
  • See references/aaron-francis-framework.md

Algorithm Alignment Without Obsession

Do This:

  • Clickable titles with genuine value
  • 30-second hook or lose viewers
  • Visual variety for retention
  • Consistent publishing schedule

Don't Do This:

  • Misleading clickbait
  • Shocked face thumbnails
  • Content you don't care about
  • Cynicism or excessive sarcasm

Remember Aaron's warning: "You can produce videos that do super well that you hate, and your insides will be hollowed out."

Key Principles for Technical Content

  1. Start with title and thumbnail, not the content
  2. Lead with problem or spectacle, not the solution
  3. Make viewers co-explorers, not students
  4. Document the journey, including failures
  5. Hide vegetables - education wrapped in entertainment
  6. Character-driven narratives - name your tools/models
  7. Visual variety - chunk into 2-3 minute segments
  8. Quality over quantity - remarkable content only
  9. Be authentically excited about your topics
  10. Build trust through consistency and honesty

Resources

For detailed implementation, see:

  • references/mark-rober-framework.md - Complete Rober methodology
  • references/aaron-francis-framework.md - Francis's technical approach
  • references/script-templates.md - 5 proven script structures
  • references/hook-examples.md - 50+ hook formulas
  • references/title-formulas.md - Title patterns by category
  • scripts/title_generator.py - Generate titles with scoring
  • scripts/hook_analyzer.py - Evaluate hook effectiveness
  • scripts/shorts_extractor.py - Extract shorts from outlines

平台分布

Codex

33.53%
按下载量换算44

Claude

30.79%
按下载量换算41

Cursor

18.38%
按下载量换算24

Gemini CLI

8.97%
按下载量换算12

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/kkoppenhaver/cc-skills --skill youtube-producer 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

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