youtube

当用户提供了YouTube链接并希望获得摘要或字幕时使用此功能。在“YouTube链接”“请求总结视频”“/youtube <URL>”等指令触发时自动激活。通过youtube-transcript-api提取字幕，并生成结构化的韩语摘要。

SKILL.md

--- frontmatter

name: "youtube"
description: "Use when the user provides a YouTube URL and wants a summary or transcript. Triggers on: YouTube links, requests to summarize a video, '/youtube <URL>'. Extracts captions via youtube-transcript-api and produces a structured Korean summary."
user_invocable: true

YouTube Video Summarizer

When to use

•User provides a YouTube URL and asks for a summary
•User types /youtube <URL>
•User asks to extract or read a YouTube video's content

Workflow

1. Extract video ID from URL

Supported formats:

•https://www.youtube.com/watch?v=VIDEO_ID
•https://youtu.be/VIDEO_ID
•https://youtube.com/shorts/VIDEO_ID

2. Fetch transcript using Python

bash

python -c "
import sys, json
from youtube_transcript_api import YouTubeTranscriptApi

video_id = 'VIDEO_ID'
ytt_api = YouTubeTranscriptApi()

# Try Korean first, then English, then any available
try:
    t = ytt_api.fetch(video_id, languages=['ko', 'en'])
except:
    t = ytt_api.fetch(video_id)

snippets = t.snippets
text = ' '.join([s.text for s in snippets])
print(text)
"

Replace VIDEO_ID with the actual video ID extracted in step 1.

3. Summarize the transcript

After fetching, provide a structured summary in Korean:

code

## [Video Title if available]

### 핵심 요약
- 3~5줄 핵심 내용

### 상세 내용
- 주요 토픽별 정리

### 키워드
- 영상에서 다룬 핵심 키워드 나열

Error handling

•If no transcript is available: inform user that the video has no captions
•If video ID is invalid: ask user to check the URL
•If package is missing: run pip install youtube-transcript-api

Notes

•Do NOT download the video itself — only fetch captions/subtitles
•Prefer Korean subtitles if available, fall back to English
•For very long transcripts (>10000 chars), chunk and summarize progressively