home / skills / zephyrwang6 / myskill / youtube-transcript-cn

youtube-transcript-cn skill

safe

This skill extracts YouTube subtitles and converts them to Chinese text, automatically translating non-Chinese captions when needed.

npx playbooks add skill zephyrwang6/myskill --skill youtube-transcript-cn

Review the files below or copy the command above to add this skill to your agents.

Files (3)

SKILL.md

2.9 KB

---
name: youtube-transcript-cn
description: |
  从 YouTube 视频链接提取字幕并转换为中文文字稿。支持自动生成字幕和手动字幕。
  使用场景：(1) 用户提供 YouTube 链接并要求提取字幕/文字稿，(2) 用户要求将 YouTube 视频内容转为文字，
  (3) 用户说"帮我把这个 YouTube 视频转成文字"或类似请求。
  触发词：YouTube 字幕、视频转文字、提取字幕、YouTube transcript、视频文字稿。
---

# YouTube 字幕提取技能

从 YouTube 视频提取字幕，输出中文文字稿。

## 快速使用

```bash
python3 scripts/get_transcript.py "https://www.youtube.com/watch?v=VIDEO_ID"
```

## 工作流程

1. **获取视频链接**：从用户消息中提取 YouTube URL
2. **运行脚本**：执行 `scripts/get_transcript.py` 提取字幕
3. **处理输出**：
   - 如果字幕是中文：直接输出
   - 如果字幕是英文：翻译为中文后输出
4. **保存结果**：根据用户需求保存为 Markdown 文件

## 脚本参数

| 参数 | 说明 | 默认值 |
|-----|------|-------|
| `url` | YouTube 视频链接或 ID | 必填 |
| `-f, --format` | 输出格式: `text`, `markdown`, `json` | `markdown` |
| `-t, --timestamps` | 包含时间戳 | 否 |
| `-l, --lang` | 首选语言（逗号分隔） | `zh,en` |
| `-o, --output` | 输出文件路径 | stdout |

## 使用示例

### 基本用法
```bash
python3 scripts/get_transcript.py "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
```

### 带时间戳输出
```bash
python3 scripts/get_transcript.py -t "https://youtu.be/VIDEO_ID"
```

### 保存到文件
```bash
python3 scripts/get_transcript.py -o output.md "https://www.youtube.com/watch?v=VIDEO_ID"
```

### JSON 格式（便于后处理）
```bash
python3 scripts/get_transcript.py -f json "VIDEO_ID"
```

## 支持的 URL 格式

- `https://www.youtube.com/watch?v=VIDEO_ID`
- `https://youtu.be/VIDEO_ID`
- `https://www.youtube.com/embed/VIDEO_ID`
- 直接使用 Video ID

## 语言优先级

脚本按以下顺序查找字幕：
1. 简体中文 (zh-Hans)
2. 繁体中文 (zh-Hant)
3. 中文 (zh)
4. 英文 (en)
5. 自动生成字幕

## 翻译处理

如果获取到的是英文字幕，需要翻译为中文：

1. 运行脚本获取英文字幕
2. 使用 Claude 内置能力翻译为流畅的中文
3. 保持原文结构和段落划分

## 错误处理

| 错误 | 原因 | 解决方案 |
|-----|------|---------|
| No captions available | 视频无字幕 | 告知用户该视频没有可用字幕 |
| Cannot extract video ID | URL 格式错误 | 请求用户提供正确的 YouTube 链接 |
| 网络超时 | 网络问题 | 重试或检查网络连接 |

## 输出格式示例

### Markdown 格式
```markdown
# 视频标题

- **Video ID**: abc123
- **语言**: 中文（自动生成）
- **链接**: https://www.youtube.com/watch?v=abc123

---

## 文字稿

这是视频的第一段内容...

这是视频的第二段内容...
```

Overview

This skill extracts subtitles from a YouTube video link and produces a Chinese transcript. It supports both uploaded/manual captions and auto-generated subtitles, and will translate English captions into natural Chinese when needed. Output can be plain text, Markdown, or JSON and can include timestamps on request.

How this skill works

The skill parses the user message to find a YouTube URL or video ID, then runs a transcript extraction routine to fetch available captions in priority: Simplified Chinese, Traditional Chinese, any Chinese, English, then auto-generated captions. If the retrieved captions are not Chinese, the text is translated into fluent Chinese while preserving paragraph breaks and timing. Finally the result is returned or saved to a file path and formatted per the requested output type.

When to use it

You have a YouTube link and need the spoken content in Chinese
You want a readable Chinese transcript from English or auto-generated captions
You need a Markdown or JSON file for editing or downstream processing
You want timestamps included for editing, quoting, or subtitles alignment

Best practices

Provide a full YouTube URL or the exact video ID to avoid extraction errors
Specify preferred output format (markdown, text, json) and whether to include timestamps
If automatic captions are low quality, request conservative translation and mark sections as ‘auto-generated’
Confirm desired language preference list (e.g., zh,en) if you expect multilingual captions
For long videos, request chunked output or file save to avoid large single-message payloads

Example use cases

Convert an English lecture video into a Chinese transcript for study notes
Generate a Markdown transcript for a podcast episode with timestamps for editing
Extract and translate captions for a tutorial video to create localized documentation
Produce JSON output with per-segment text and timestamps for automated processing
Save a video transcript to a file for archival or content repurposing

FAQ

What if the video has no captions?

I will inform you that no captions are available and suggest enabling auto-generated subtitles or providing a different link.

Can I get timestamps in the transcript?

Yes — you can request timestamps; the script can include per-segment timing in the chosen output format.