重要前提
安装AI Skills的关键前提是:必须科学上网,且开启TUN模式,这一点至关重要,直接决定安装能否顺利完成,在此郑重提醒三遍:科学上网,科学上网,科学上网。查看完整安装教程 →
plan-video by kenneth-liao/ai-launchpad-marketplace
npx skills add https://github.com/kenneth-liao/ai-launchpad-marketplace --skill plan-video此编排器按顺序调用基础技能,以生成完整的视频规划:研究过的标题、缩略图概念、钩子策略和内容大纲。它管理步骤之间的用户选择工作流,但将所有内容生成委托给专业化的技能。
核心原则:这是一个轻量级的编排器。所有内容生成都通过基础技能完成——那里存放着经过验证的模式、公式和质量逻辑。此技能仅管理工作流顺序、用户决策和规划文件组装。
下面的每个步骤都会调用一个拥有其自身引用文件的基础技能。当你调用 creator-stack:title 时,它会从其自身目录加载 references/youtube-title-formulas.md。当你调用 creator-stack:thumbnail 时,它会从其自身目录加载 references/thumbnail-formulas.md。你无需管理这些路径——只需调用技能并提供上下文。
所有剧集文件都位于 ./youtube/episode/[episode_number]_[topic_short_name]/ 下。
research.md广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
plan.md如果存在本地频道上下文文件(例如 ./youtube/channel.md 或类似文件),在工作流开始时加载它,以便为下游技能提供频道特定的上下文。
当从此编排器调用 creator-stack:research 时,YouTube 分析 MCP 工具可用于数据收集:
search_videos — 通过关键词、频道或条件查找视频get_video_details — 视频统计数据、观看量、点赞、评论、发布日期get_channel_details — 频道元数据、订阅者数量、视频数量get_video_comments — 评论文本和情感数据get_video_transcript — 完整视频转录get_related_videos — 与特定视频相关的视频get_trending_videos — 当前热门视频get_video_enhanced_transcript — 带时间戳的增强转录YouTube 研究员代理(在此插件的 agents/youtube-researcher.md 中定义)也可用于结构化数据收集任务。
在以下情况下使用此技能:
必须先完成视频的研究。要么:
./youtube/episode/[episode_number]_[topic_short_name]/research.md,或者creator-stack:research 进行研究(YouTube MCP 工具可用)按顺序执行以下所有步骤。系统地跟踪进度。
读取此剧集的研究文件:
./youtube/episode/[episode_number]_[topic_short_name]/research.md如果研究不存在,首先调用 creator-stack:research。YouTube MCP 工具在此编排器上下文中可用,使研究技能能够收集 YouTube 特定数据。
视频规划应在每个步骤中纳入研究发现——研究为每个推荐提供依据。
在以下位置创建新的规划文件:./youtube/episode/[episode_number]_[topic_short_name]/plan.md
如果规划文件已存在,请读取它以了解到目前为止已完成的工作,并从中断处继续。
使用此模板初始化规划文件:
# [Episode_Number]: [Topic] - Video Plan
## Research Summary
[Summary of research insights]
## Titles
[3 title options with rationale and star ratings]
## Thumbnails
[2 thumbnail concepts per title with rationale and star ratings]
## Hooks
[3 hook options for selected title+thumbnail with rationale and star ratings]
## High-Level Content Outline
[Section structure with key points]
## Final Plan
[Selected title, 3 AB test thumbnails, selected hook, outline]
在完成规划步骤的过程中更新此文档。
调用 creator-stack:title 生成 3 个优化标题。
在规划文件中记录所有标题选项,包括:
在继续缩略图之前完成标题生成——缩略图需要与标题互补。
调用 creator-stack:thumbnail 为每个标题生成 2 个缩略图概念。这些应仅为概念描述,而非实际图像。
在规划文件中记录所有缩略图概念,包括:
向用户呈现所有标题/缩略图组合,以及你推荐的 3 个最佳标题+缩略图配对。
请用户选择他们偏好的一个标题+缩略图配对以继续。
更新规划文件中的标题和缩略图部分,在所选选项旁用(用户选择)标记指示用户的选择。
始终呈现所有选项,以便用户做出明智的决定——这正是生成多个选项的全部意义。
调用 creator-stack:hook 生成 3 个优化留存率的钩子,仅针对用户选择的标题+缩略图配对。
在规划文件中记录所有钩子策略,包括:
向用户呈现所有钩子选项,并请用户选择他们偏好的钩子策略。
更新规划文件中的钩子部分,指示用户的选择。
规划现在应包含用户选择的标题、缩略图和钩子组合。
调用 creator-stack:write 创建高层内容大纲。
创建并记录战略路线图:
保持大纲的高层性——仅包含结构和要点,无需详细脚本。专注于从观众视角来看需要涵盖的重要内容,而不是假设特定的内容。
既然用户已经选择了他们偏好的标题、缩略图和钩子,现在可以完成规划。
调用 creator-stack:thumbnail 生成 3 个用于 AB 测试的缩略图选项。这些应该是使用 creator-stack:nanobanana 生成的实际图像,而不仅仅是概念。第一个缩略图应基于用户选择的缩略图概念。另外 2 个应测试与第一个缩略图不同的视觉风格。
使用完整的最终选择更新规划文件中的最终规划部分:
## Final Plan
- **Title**: [Selected title]
- **Thumbnails**:
- 
Thumbnail A Description
- 
Thumbnail B Description
- 
Thumbnail C Description
- **Hook**: [Selected hook strategy]
- **Rationale**: [Why this combination works]
最终规划始终包含 3 个用于 AB 测试的缩略图——这是你了解什么对你的受众有效的方式。
每个基础技能拥有其自身的引用文件并自动加载它们:
| 内容 | 技能 | 引用(由技能加载) |
|---|---|---|
| 标题 | creator-stack:title | youtube-title-formulas.md |
| 缩略图 | creator-stack:thumbnail | thumbnail-formulas.md |
| 钩子 | creator-stack:hook | youtube-hooks.md |
| 大纲 | creator-stack:write | youtube-script.md |
| 语音 | creator-stack:voice | 由写作技能自动调用 |
在最终确定之前,验证标题/缩略图/钩子是否协同工作:
在最终确定规划前验证完成情况:
creator-stack:title 已调用 — 生成 3 个标题选项creator-stack:thumbnail 已调用 — 每个标题生成 2 个缩略图概念creator-stack:hook 已调用 — 生成 3 个钩子策略creator-stack:write 已调用 — 创建高层内容大纲creator-stack:thumbnail 已调用 — 生成 3 个 AB 测试缩略图场景:用户请求规划关于“使用记忆构建 AI 代理”的视频(研究已完成)
./youtube/episode/18_ai_agents_with_memory/research.md,提取关键洞察./youtube/episode/18_ai_agents_with_memory/plan.md 创建规划文件creator-stack:title — 3 个专注于实际实现的标题选项creator-stack:thumbnail — 每个标题 2 个概念(共 6 个)creator-stack:hook — 为选定配对生成 3 个钩子策略creator-stack:write — 高层内容大纲creator-stack:thumbnail — 为最终规划生成 3 个 AB 测试缩略图结果:生产就绪的规划,包含供用户选择的多个选项,所有选项均有研究支持,并使用基础技能中经过验证的模式生成。
每周安装量
55
仓库
GitHub 星标
124
首次出现
2026年2月19日
安全审计
安装于
cursor55
opencode55
gemini-cli54
kimi-cli54
github-copilot54
amp54
This orchestrator sequences foundation skills to produce a complete video plan: researched titles, thumbnail concepts, hook strategies, and a content outline. It manages the user selection workflow between steps but delegates all content generation to specialized skills.
Core Principle : This is a thin orchestrator. All content generation goes through foundation skills — that's where the proven patterns, formulas, and quality logic live. This skill manages workflow sequence, user decisions, and plan file assembly only.
Each step below invokes a foundation skill that owns its own reference files. When you invoke creator-stack:title, it loads references/youtube-title-formulas.md from its own directory. When you invoke creator-stack:thumbnail, it loads references/thumbnail-formulas.md from its own directory. You don't need to manage these paths — just invoke the skill and provide context.
All episode files live under ./youtube/episode/[episode_number]_[topic_short_name]/.
research.md in the episode directoryplan.md in the episode directoryIf a local channel context file exists (e.g., ./youtube/channel.md or similar), load it at the start of the workflow to provide channel-specific context to downstream skills.
When creator-stack:research is invoked from this orchestrator, the YouTube Analytics MCP tools are available for data gathering:
search_videos — find videos by keyword, channel, or criteriaget_video_details — video stats, views, likes, comments, publish dateget_channel_details — channel metadata, subscriber count, video countget_video_comments — comment text and sentiment dataget_video_transcript — full video transcriptget_related_videos — videos related to a specific videoget_trending_videos — currently trending videosget_video_enhanced_transcript — enhanced transcript with timestampsThe YouTube Researcher agent (defined in this plugin's agents/youtube-researcher.md) can also be used for structured data gathering tasks.
Use this skill when:
Research for the video must be completed first. Either:
./youtube/episode/[episode_number]_[topic_short_name]/research.md, ORcreator-stack:research to conduct research first (YouTube MCP tools are available)Execute all steps below in order. Track progress systematically.
Read the research file for this episode:
./youtube/episode/[episode_number]_[topic_short_name]/research.mdIf research does not exist, invoke creator-stack:research first. The YouTube MCP tools are available from this orchestrator context, enabling the research skill to gather YouTube-specific data.
The video plan should incorporate research findings at each step — research grounds every recommendation.
Create a new plan file at: ./youtube/episode/[episode_number]_[topic_short_name]/plan.md
If a plan file already exists, read it to understand what has been done so far and continue from there.
Initialize the plan file with this template:
# [Episode_Number]: [Topic] - Video Plan
## Research Summary
[Summary of research insights]
## Titles
[3 title options with rationale and star ratings]
## Thumbnails
[2 thumbnail concepts per title with rationale and star ratings]
## Hooks
[3 hook options for selected title+thumbnail with rationale and star ratings]
## High-Level Content Outline
[Section structure with key points]
## Final Plan
[Selected title, 3 AB test thumbnails, selected hook, outline]
Update this document as you progress through the planning steps.
Invoke creator-stack:title to generate 3 optimized titles.
Document all title options in the plan file including:
Complete title generation before proceeding to thumbnails — thumbnails need to complement the titles.
Invoke creator-stack:thumbnail to generate 2 thumbnail concepts for each title. These should be concept descriptions only, not actual images yet.
Document all thumbnail concepts in the plan file including:
Present all title/thumbnail combinations to the user, along with your top 3 recommended title + thumbnail pairings.
Ask the user to select their one preferred title + thumbnail pairing to proceed with.
Update both the title and thumbnail sections in the plan file to indicate the user's selection with a (User Selection) marker next to the selected options.
Always present all options so the user can make an informed decision — that's the whole point of generating multiple options.
Invoke creator-stack:hook to generate 3 retention-optimized hooks, only for the user's selected title + thumbnail pairing.
Document all hook strategies in the plan file including:
Present all hook options to the user and ask the user to select their preferred hook strategy.
Update the hook section in the plan file to indicate the user's selection.
The plan should now contain a user-selected title, thumbnail, and hook combination.
Invoke creator-stack:write to create a high-level content outline.
Create and document the strategic roadmap:
Keep the outline high-level — structure and key points only, no detailed scripts. Focus on what is important to cover from the viewer's perspective rather than assuming specific content.
Now that the user has selected their preferred title, thumbnail, and hook, finalize the plan.
Invoke creator-stack:thumbnail to generate 3 thumbnail options for AB testing. These should be actual images generated with creator-stack:nanobanana, not just concepts. The first thumbnail should be based on the user's selected thumbnail concept. The other 2 should test different visual styles of the first thumbnail.
Update the final plan section in the plan file with the complete final selections:
## Final Plan
- **Title**: [Selected title]
- **Thumbnails**:
- 
Thumbnail A Description
- 
Thumbnail B Description
- 
Thumbnail C Description
- **Hook**: [Selected hook strategy]
- **Rationale**: [Why this combination works]
A final plan always includes 3 thumbnails to AB test — that's how you learn what works for your audience.
Each foundation skill owns its own reference files and loads them automatically:
| Content | Skill | Reference (loaded by the skill) |
|---|---|---|
| Titles | creator-stack:title | youtube-title-formulas.md |
| Thumbnails | creator-stack:thumbnail | thumbnail-formulas.md |
| Hooks | creator-stack:hook | youtube-hooks.md |
| Outline |
Before finalizing, verify that title/thumbnail/hook work together:
Verify completion before finalizing plan:
creator-stack:title invoked — 3 title options generatedcreator-stack:thumbnail invoked — 2 thumbnail concepts per titlecreator-stack:hook invoked — 3 hook strategies generatedcreator-stack:write invoked — high-level content outline createdcreator-stack:thumbnail invoked — 3 AB testing thumbnails generatedScenario : User requests plan for video about "Building AI agents with memory" (research already complete)
./youtube/episode/18_ai_agents_with_memory/research.md, extract key insights./youtube/episode/18_ai_agents_with_memory/plan.mdcreator-stack:title — 3 title options focused on practical implementationcreator-stack:thumbnail — 2 concepts per title (6 total)creator-stack:hook — 3 hook strategies for selected pairingcreator-stack:write — high-level content outlinecreator-stack:thumbnail — 3 AB testing thumbnails for final planResult : Production-ready plan with multiple options for user selection, all backed by research and generated using proven patterns from foundation skills.
Weekly Installs
55
Repository
GitHub Stars
124
First Seen
Feb 19, 2026
Security Audits
Gen Agent Trust HubPassSocketPassSnykWarn
Installed on
cursor55
opencode55
gemini-cli54
kimi-cli54
github-copilot54
amp54
专业SEO审计工具:全面网站诊断、技术SEO优化与页面分析指南
73,800 周安装
creator-stack:writeyoutube-script.md |
| Voice | creator-stack:voice | Invoked automatically by writing skill |