studio-setup

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Studio Setup

工作室设置

Ten minutes, once. At the end the creator has:

the tools the other skills call, installed and checked,
```
video-agent/voice.md
```
: who they are, how they talk, what they believe, what they can prove. Every script is written from it,
the editing studio ready, so the first edit starts right away,
permission rules, so Claude can run ffmpeg, yt-dlp and Remotion without asking each time,
if they want: their accounts connected through Treg (audit and publishing), and the keys for translation and images.

一次只需十分钟。完成后，创作者将拥有：

已安装并检查其他 skill 所调用的工具，
```
video-agent/voice.md
```
：他们是谁、如何说话、相信什么以及能证明什么。每个脚本都据此编写，
已准备就绪的剪辑工作室，让第一次剪辑可以立即开始，
权限规则，让 Claude 可以运行 ffmpeg、yt-dlp 和 Remotion，而不必每次询问，
如有需要：通过 Treg 连接他们的账户（数据分析和发布），以及用于翻译和图像生成的 key。

How to run it

运行方式

Talk in the creator's language. Simple words, no jargon.
One question per message. Offer choices when you can, recommended first.
Never invent anything about them: a fact you don't have becomes a question or a
```
{{placeholder}}
```
.

使用创作者的语言交流。使用简单词汇，不使用行话。
每条消息只问一个问题。尽可能提供选项，并优先给出推荐项。
绝不编造关于他们的任何信息：没有依据的事实应转化为问题或
```
{{placeholder}}
```
。

1. Tools

1. 工具

Run the checks and show one line per tool (✓ version, or ✗ missing):

bash

node -v
ffmpeg -version
yt-dlp --version
deno --version
git --version

Tool	Used by	macOS	Windows	Linux
Node 18+	short-editing (Remotion)	`brew install node`	`winget install OpenJS.NodeJS.LTS`	nodejs.org or your package manager
ffmpeg	short-editing, steal, clips, short-translate, publish	`brew install ffmpeg`	`winget install Gyan.FFmpeg`	`sudo apt install ffmpeg`
yt-dlp	studio-script viral, steal, clips, voice.md from videos	`brew install yt-dlp`	`winget install yt-dlp.yt-dlp`	`pipx install "yt-dlp[default]"`
deno	yt-dlp on YouTube (it needs a JavaScript runtime)	comes with yt-dlp	`winget install DenoLand.Deno`	`curl -fsSL https://deno.land/install.sh \| sh`
git, build tools	short-editing (gets the Remotion template, builds whisper.cpp)	`xcode-select --install`	`winget install Git.Git`	`sudo apt install git build-essential`
Codex (optional)	codex-images	`npm i -g @openai/codex` , then `codex login`	same	same

Give the commands for what is missing and let them run them (or run them yourself if they say so). Scripting-only users (studio-script, its

viral

mode included) only need yt-dlp and deno. Codex is only for

/codex-images

and needs a ChatGPT plan; skip it if they don't want images.

Windows: after
```
winget
```
, close and reopen the terminal (and Claude Code), or the new commands are not found.
YouTube changes often. When yt-dlp fails on a link that plays in the browser, update it first (
```
brew upgrade yt-dlp
```
,
```
winget upgrade yt-dlp.yt-dlp
```
,
```
pipx upgrade yt-dlp
```
).

运行检查，并为每个工具显示一行结果（✓ 版本，或 ✗ 缺失）：

bash

node -v
ffmpeg -version
yt-dlp --version
deno --version
git --version

工具	用于	macOS	Windows	Linux
Node 18+	short-editing（Remotion）	`brew install node`	`winget install OpenJS.NodeJS.LTS`	nodejs.org 或你的包管理器
ffmpeg	short-editing、steal、clips、short-translate、publish	`brew install ffmpeg`	`winget install Gyan.FFmpeg`	`sudo apt install ffmpeg`
yt-dlp	studio-script viral、steal、clips、从视频生成 voice.md	`brew install yt-dlp`	`winget install yt-dlp.yt-dlp`	`pipx install "yt-dlp[default]"`
deno	在 YouTube 上运行 yt-dlp（需要 JavaScript 运行时）	随 yt-dlp 一同安装	`winget install DenoLand.Deno`	`curl -fsSL https://deno.land/install.sh \| sh`
git、构建工具	short-editing（获取 Remotion 模板、构建 whisper.cpp）	`xcode-select --install`	`winget install Git.Git`	`sudo apt install git build-essential`
Codex（可选）	codex-images	`npm i -g @openai/codex` ，然后执行 `codex login`	相同	相同

针对缺失项给出命令，让他们自行运行；如果他们要求，也可以由你代为运行。仅使用脚本的用户（包括 studio-script 及其

viral

模式）只需要 yt-dlp 和 deno。Codex 仅用于

/codex-images

，并且需要 ChatGPT 套餐；如果他们不需要生成图像，则跳过。

Windows：使用
```
winget
```
后，关闭并重新打开终端（以及 Claude Code），否则系统找不到新命令。
YouTube 经常变化。如果 yt-dlp 无法处理某个在浏览器中可以播放的链接，请先更新它（
```
brew upgrade yt-dlp
```
、
```
winget upgrade yt-dlp.yt-dlp
```
、
```
pipx upgrade yt-dlp
```
）。

2. The folder

2. 文件夹

Create it at the root of the current project:

text

video-agent/
├── voice.md      who they are and how they talk (step 4)
├── swipe.md      hooks that worked in their niche (`/studio-script viral`)
├── audit.md      what works on their own Instagram (insta-audit)
├── audits/       full Instagram reports, one folder per account and run (insta-audit @handle)
├── log.md        one line per script written, so nothing repeats
├── scripts/      one file per week (studio-script)
├── takes/        raw recordings (phone, camera, webcam, screen)
├── styles/       styles copied from other creators (steal): <name>.md + its stills, sfx, stickers
├── edits/        one folder per edited video (short-editing): final.mp4, post.md, cover.jpg, published.md
├── repurpose/    one plan per long video (clips)
├── images/       generated images not tied to a video (codex-images)
├── studio/       the Remotion project (step 3)
├── .env          their API keys, never shared (step 6)
└── .tmp/         scratch files, deleted after each run

log.md

starts with

# Script log

swipe.md

with

# Swipe file

If the project is in git, add these lines to its

.gitignore

(videos are heavy; the text files stay tracked):

gitignore

video-agent/takes/
video-agent/edits/**/*.mp4
video-agent/studio/public/
video-agent/studio/out/
video-agent/studio/node_modules/
video-agent/studio/whisper.cpp/
video-agent/.tmp/
video-agent/.env

video-agent/.env

holds keys: it must be ignored even if they skip the rest.

在当前项目根目录创建：

text

video-agent/
├── voice.md      who they are and how they talk (step 4)
├── swipe.md      hooks that worked in their niche (`/studio-script viral`)
├── audit.md      what works on their own Instagram (insta-audit)
├── audits/       full Instagram reports, one folder per account and run (insta-audit @handle)
├── log.md        one line per script written, so nothing repeats
├── scripts/      one file per week (studio-script)
├── takes/        raw recordings (phone, camera, webcam, screen)
├── styles/       styles copied from other creators (steal): <name>.md + its stills, sfx, stickers
├── edits/        one folder per edited video (short-editing): final.mp4, post.md, cover.jpg, published.md
├── repurpose/    one plan per long video (clips)
├── images/       generated images not tied to a video (codex-images)
├── studio/       the Remotion project (step 3)
├── .env          their API keys, never shared (step 6)
└── .tmp/         scratch files, deleted after each run

log.md

以

# Script log

开头，

swipe.md

以

# Swipe file

开头。

如果项目使用 git，请将以下行添加到其

.gitignore

中（视频文件很大；文本文件仍保留在版本控制中）：

gitignore

video-agent/takes/
video-agent/edits/**/*.mp4
video-agent/studio/public/
video-agent/studio/out/
video-agent/studio/node_modules/
video-agent/studio/whisper.cpp/
video-agent/.tmp/
video-agent/.env

video-agent/.env

用于存放 key：即使创作者跳过其余所有设置，也必须忽略此文件。

3. The studio

3. 工作室

The editing studio (Remotion, whisper.cpp for the captions, a speech model) takes 5 to 15 minutes and about 2 GB to set up. Ask: now (recommended if they will edit soon, the first edit then starts right away), or later (short-editing does it the first time it runs).

Now: run step 2 of

short-editing

("The studio"), including the

Short

and

Wide

compositions and the

npx tsc --noEmit

check. Say what is downloading and how big before starting. Only for short-editing, clips and short-translate; scripting-only users skip it.

剪辑工作室（Remotion、用于生成字幕的 whisper.cpp 以及一个语音模型）的设置需要 5 到 15 分钟和约 2 GB 空间。询问：现在还是稍后？如果他们很快就要剪辑，建议选择现在，这样第一次剪辑可以立即开始；如果选择稍后，

short-editing

会在首次运行时完成设置。

现在：运行

short-editing

的第 2 步（“The studio”），包括创建

Short

和

Wide

compositions，并执行

npx tsc --noEmit

检查。开始前说明正在下载的内容及其大小。仅 short-editing、clips 和 short-translate 需要执行；仅使用脚本的用户跳过。

4. voice.md

First the two settings at the top of the file: the language they record in, and their brand color (the accent of captions and cards; default

#FFD400

, a yellow that reads on any footage). Then ask how to build the rest:

From their videos (recommended when they have posted before): 3 to 5 links of their best talking videos (YouTube, Shorts, TikTok, Instagram reels). Pull the transcripts (
```
<lang>
```
= their language code:
```
en
```
,
```
fr
```
,
```
es
```
):
bash
```
yt-dlp --skip-download --write-auto-subs --write-subs --sub-langs "<lang>.*" --sub-format vtt -o "video-agent/.tmp/%(id)s.%(ext)s" "<url>"
```
Read them and fill the template from what they actually say: their real phrases, their pace (words ÷ minutes of speech), the words they repeat, the way they open and close. Then ask only what the transcripts can't tell (positions, proof, off limits, the ask). Delete
```
video-agent/.tmp/
```
after. YouTube links almost always give a transcript. TikTok sometimes does, Instagram almost never: for those, ask for a YouTube link or switch to the interview.
From an interview: ask the template's questions one at a time, in plain words ("What do you believe that most people in your niche get wrong?").

Either way, ask them to paste 3 short texts they wrote themselves, before any AI tool (a caption, a message, an email). They go in "Real samples" and teach the scripts their rhythm.

Write

video-agent/voice.md

from references/voice-template.md. Then read it back to them in 5 lines and ask what is wrong. A voice file is never finished: say they can run

/studio-setup voice

any time to update it.

先设置文件顶部的两项内容：他们录制的语言，以及品牌色（字幕和卡片的强调色；默认为

#FFD400

，这种黄色在任何画面上都清晰可见）。然后询问如何构建其余内容：

从他们的视频构建（此前发布过内容时推荐）：提供 3 到 5 个表现最佳的出镜讲话视频链接（YouTube、Shorts、TikTok、Instagram Reels）。获取转录文本（
```
<lang>
```
为他们的语言代码：
```
en
```
、
```
fr
```
、
```
es
```
）：
bash
```
yt-dlp --skip-download --write-auto-subs --write-subs --sub-langs "<lang>.*" --sub-format vtt -o "video-agent/.tmp/%(id)s.%(ext)s" "<url>"
```
阅读这些转录文本，并根据他们实际所说的内容填写模板：他们真实的表达方式、语速（词数 ÷ 讲话分钟数）、经常重复的词，以及开场和收尾方式。然后只询问转录文本无法提供的信息（立场、证据、不可提及内容、最终诉求）。完成后删除
```
video-agent/.tmp/
```
。 YouTube 链接几乎总能提供转录文本。TikTok 有时提供，而 Instagram 几乎从不提供；对于这些平台，请要求他们提供 YouTube 链接，或改用访谈方式。
通过访谈构建：用平实语言逐个询问模板中的问题（“在你的细分领域，大多数人都误解了什么？”）。

无论采用哪种方式，都请他们提供在使用任何 AI 工具之前亲自撰写的 3 段短文本（一条标题、一条消息或一封电子邮件）。这些内容会放入“Real samples”，用于教脚本掌握他们的表达节奏。

根据 references/voice-template.md 编写

video-agent/voice.md

。然后用 5 行内容概括并读给他们听，询问有哪些问题。声音档案永远没有最终版本：告诉他们可以随时运行

/studio-setup voice

进行更新。

5. Permissions

5. 权限

The other skills run the same few commands again and again. Pre-approve them for this project. Show the rules, ask, then merge them into

.claude/settings.json

. Create the file if it is missing. Keep every key and rule already there, and never remove anything.

json

{
  "permissions": {
    "allow": [
      "Bash(yt-dlp *)",
      "Bash(ffmpeg *)",
      "Bash(ffprobe *)",
      "Bash(npx remotion *)",
      "Bash(node sub.mjs *)"
    ]
  }
}

Only these. Never add
```
Bash(*)
```
,
```
Bash(rm *)
```
,
```
Bash(curl *)
```
or a rule for a whole interpreter (
```
Bash(node *)
```
,
```
Bash(python *)
```
).
Rules in the shared
```
settings.json
```
apply once the folder is trusted. A creator who keeps this project in git and doesn't want to share the rules can use
```
.claude/settings.local.json
```
instead (same format, personal).
Each skill also lists its own commands in its
```
allowed-tools
```
, which covers the turn the skill runs in. These rules cover every other turn.

其他 skill 会反复运行相同的几条命令。为该项目预先批准这些命令。先展示规则并征求同意，然后将其合并到

.claude/settings.json

中。如果文件不存在，则创建。保留其中已有的每个 key 和规则，绝不删除任何内容。

json

{
  "permissions": {
    "allow": [
      "Bash(yt-dlp *)",
      "Bash(ffmpeg *)",
      "Bash(ffprobe *)",
      "Bash(npx remotion *)",
      "Bash(node sub.mjs *)"
    ]
  }
}

仅限这些规则。绝不添加
```
Bash(*)
```
、
```
Bash(rm *)
```
、
```
Bash(curl *)
```
，也不要添加针对整个解释器的规则（
```
Bash(node *)
```
、
```
Bash(python *)
```
）。
共享
```
settings.json
```
中的规则会在文件夹被信任后生效。如果创作者希望将项目保留在 git 中且不想共享这些规则，可以改用
```
.claude/settings.local.json
```
（格式相同，属于个人设置）。
每个 skill 还会在其
```
allowed-tools
```
中列出自己的命令，这会覆盖该 skill 运行所在的这一轮对话。上述规则则覆盖其他所有轮次。

6. Accounts and keys (optional)

6. 账户和 key（可选）

Three things, each only if they want the skill that needs it. Ask once, in one message:

For	What	Cost
`/insta-audit` and `/publish`	Treg, with their Instagram, TikTok and YouTube connected	free on their own accounts
`/short-translate` (and `/steal` 's music level and sound effects, optional)	an ElevenLabs API key	their ElevenLabs plan
`/codex-images`	Codex logged in with their ChatGPT account	their ChatGPT plan

ElevenLabs: they create a key at elevenlabs.io/app/settings/api-keys and paste it themselves into

video-agent/.env

ELEVENLABS_API_KEY=...

(create the empty file and open it for them). Never ask for a key in the chat. Check it is there without printing it:

grep -c ELEVENLABS_API_KEY video-agent/.env

Codex:

codex --version

, then they run

codex login

(browser, "Sign in with ChatGPT").

共有三件事，每项仅在他们需要相应 skill 时才设置。在一条消息中一次性询问：

用途	所需内容	成本
`/insta-audit` 和 `/publish`	Treg，连接其 Instagram、TikTok 和 YouTube	在自有账户上免费
`/short-translate` （以及 `/steal` 的音乐音量和音效，可选）	ElevenLabs API key	其 ElevenLabs 套餐
`/codex-images`	使用 ChatGPT 账户登录 Codex	其 ChatGPT 套餐

ElevenLabs：他们可以在 elevenlabs.io/app/settings/api-keys 创建 key，然后自行粘贴到

video-agent/.env

中，格式为

ELEVENLABS_API_KEY=...

（创建空文件并为其打开）。绝不要在聊天中索要 key。通过以下命令检查 key 是否存在，但不要打印它：

grep -c ELEVENLABS_API_KEY video-agent/.env

。

Codex：执行

codex --version

，然后让他们运行

codex login

（在浏览器中选择“Sign in with ChatGPT”）。

Treg

yt-dlp reads YouTube for free. TikTok and Instagram hide their numbers from scripts, a creator's own stats sit behind their login, and posting needs their accounts. Treg fills that gap: their own accounts connected once (stats and publishing, free), other creators' public numbers for about $0.001 a call. New verified accounts get $1.00 of credit once, enough for hundreds of calls.

Two ways to plug it in:

CLI, macOS and Linux. It needs uv (recommended, it fetches the right Python), or pipx or pip with Python 3.12 or 3.13. There is no Windows installer.
bash
```
curl -fsSL https://treg.to/install.sh | sh
treg login
```
```
treg login
```
opens the browser (GitHub, Google or an email code).

MCP, any OS, nothing to install:

claude mcp add --transport http treg https://treg.to/mcp/

, then

/mcp

to sign in. Each

treg catalog search

catalog get

call

and

balance

below becomes the MCP tool

catalog_search

catalog_get

call

balance

Their own accounts (connecting needs the CLI).

--capability post

asks for the rights to read their stats and to post; posting still waits for their yes on each video:

bash

treg connections connect --provider instagram --capability post
treg connections connect --provider tiktok --capability post
treg connections connect --provider youtube --capability post
treg connections ls

Each

connect

opens the platform's own login; they approve it there. Connect only the platforms they post on. Instagram needs a professional account (Creator or Business, a free switch in the app: Profile → menu → Account type and tools). Treg prints tips on stderr: add

2>/dev/null

when you read its JSON.

Instagram connected: offer

/insta-audit

right after, it is the best first look at what works for them.

How every Treg call works, the rule for all the skills:

```
treg catalog search "<what you want to do>"
```
: describe the job, not a vendor ("list a TikTok account's videos with view counts").
```
treg catalog get <id>
```
: the parameters and the price. Say the price before calling. Their own connected accounts are marked free.
```
treg call <id> --query key=value
```
. A failed call is not charged.
```
treg balance
```
shows what is left; a 402 error means the balance is empty.

Permissions: add only the free, read-only commands, so every paid

treg call

still asks first:

json

"Bash(treg catalog *)",
"Bash(treg balance)",
"Bash(treg connections ls)"

yt-dlp 可以免费读取 YouTube。TikTok 和 Instagram 会向脚本隐藏数据，创作者自己的统计信息位于登录后界面中，而发布内容需要连接账户。Treg 填补了这一空白：一次连接他们自己的账户即可进行数据分析和发布，免费；查询其他创作者的公开数据时，每次调用约需 $0.001。新完成验证的账户可一次性获得 $1.00 额度，足以支持数百次调用。

有两种接入方式：

CLI，支持 macOS 和 Linux。它需要 uv（推荐，它会获取合适的 Python），或 pipx，或带 Python 3.12 或 3.13 的 pip。Windows 没有安装程序。
bash
```
curl -fsSL https://treg.to/install.sh | sh
treg login
```
```
treg login
```
会打开浏览器（GitHub、Google 或电子邮件验证码）。

MCP，支持任何 OS，无需安装：执行

claude mcp add --transport http treg https://treg.to/mcp/

，然后使用

/mcp

treg catalog search

、

catalog get

、

call

和

balance

都会成为对应的 MCP 工具

catalog_search

、

catalog_get

、

call

或

balance

。

他们自己的账户（连接需要使用 CLI）。

--capability post

请求读取统计信息和发布的权限；即使如此，每次发布视频仍需他们确认：

bash

treg connections connect --provider instagram --capability post
treg connections connect --provider tiktok --capability post
treg connections connect --provider youtube --capability post
treg connections ls

每个

connect

都会打开相应平台的登录页面；他们在那里批准授权。只连接实际发布内容的平台。Instagram 需要专业账户（Creator 或 Business，可在 App 中免费切换：Profile → menu → Account type and tools）。Treg 会将提示输出到 stderr；读取 JSON 时可添加

2>/dev/null

。

连接 Instagram 后，立即提供

/insta-audit

；这是了解其账号上哪些内容有效的最佳第一步。

每次 Treg 调用的工作方式，这是所有 skill 共同遵循的规则：

```
treg catalog search "<what you want to do>"
```
：描述任务，而不是指定供应商（“列出某个 TikTok 账户的视频及观看次数”）。
```
treg catalog get <id>
```
：查看参数及价格。调用前先说明价格。已连接的自有账户会标记为免费。
```
treg call <id> --query key=value
```
。调用失败不收费。
```
treg balance
```
显示剩余额度；402 错误表示余额已耗尽。

权限：仅添加免费、只读的命令，使每次付费的

treg call

仍会先征求确认：

json

"Bash(treg catalog *)",
"Bash(treg balance)",
"Bash(treg connections ls)"

7. Wrap up

7. 收尾

Show what is ready (✓ or skipped, one line each) and the path forward. They can also just type

/editing <what they want>

: it picks the right skill.

```
/insta-audit
```
: what works on their own Instagram (fills audit.md).
```
/studio-script viral
```
: what works in their niche (fills swipe.md; once a month).
```
/studio-script
```
: this week's scripts, hooks first, humanized.
Record them: phone, camera or webcam, one take per script, saved in
```
video-agent/takes/<slug>.mp4
```
.
```
/short-editing
```
: a short (vertical) or a full YouTube video (horizontal).
```
/steal
```
: a creator's reel they love into a style
```
short-editing
```
can use.
```
/clips
```
: a long video into finished shorts.
```
/short-translate
```
: a take in another language, in their voice.
```
/codex-images
```
: thumbnails and covers.
```
/publish
```
: YouTube, Instagram and TikTok.

展示已经准备就绪的项目（✓ 或已跳过，每项一行）以及后续路径。他们也可以直接输入

/editing <他们想要的内容>

；它会选择合适的 skill。

```
/insta-audit
```
：了解他们自己的 Instagram 上哪些内容有效（填充 audit.md）。
```
/studio-script viral
```
：了解其细分领域中哪些内容有效（填充 swipe.md；每月一次）。
```
/studio-script
```
：生成本周的脚本，先提供 hooks，并使其自然化。
录制脚本：使用手机、相机或摄像头，每个脚本录制一条，保存到
```
video-agent/takes/<slug>.mp4
```
。
```
/short-editing
```
：制作一条竖屏短视频或一条横屏 YouTube 视频。
```
/steal
```
：将喜欢的创作者的 Reel 转换为
```
short-editing
```
可使用的风格。
```
/clips
```
：将长视频转换为成品短视频。
```
/short-translate
```
：以他们自己的声音将一条素材翻译为另一种语言。
```
/codex-images
```
：生成缩略图和封面。
```
/publish
```
：发布到 YouTube、Instagram 和 TikTok。

Later

后续操作

/studio-setup <part>

jumps to one part:

tools

studio

voice

(rebuild or update voice.md),

permissions

treg

keys

/studio-setup <part>

可以直接跳转到某一部分：

tools

、

studio

、

voice

（重建或更新 voice.md）、

permissions

、

treg

、

keys

。