Youtube Transcript

Name: Youtube Transcript
Author: aiskillstore

aiskillstore/marketplace

1 installs
404 repo stars
Updated August 5, 2026
aiskillstore/marketplace

youtube-transcript is a skill that downloads and processes YouTube video subtitles with yt-dlp to produce clean transcripts and summaries.

About

youtube-transcript downloads and processes YouTube video transcripts using yt-dlp. A developer uses it to extract subtitles, convert VTT captions to clean text, and generate summaries or chapter breakdowns from video content. It prefers manual subtitles, falls back to auto-generated ones, and can transcribe audio with Whisper when no subtitles exist. The documentation is written in Korean.

Downloads YouTube subtitles with yt-dlp
Prefers manual subs, falls back to auto and then Whisper
Converts VTT to clean text and dedupes lines

Youtube Transcript by the numbers

1 all-time installs (skills.sh)
Ranked #1,980 of 2,715 Automation & Workflows skills by installs in the Skillselion catalog
Data as of Aug 5, 2026 (Skillselion catalog sync)

At a glance

youtube-transcript capabilities & compatibility

Capabilities: transcription · research
Use cases: transcription · research
Platforms: macOS · Linux

From the docs

What youtube-transcript says it does

Download and process YouTube video transcripts using yt-dlp.

SKILL.md

yt-dlp --write-sub --sub-lang en --skip-download "VIDEO_URL"

SKILL.md

자동 생성 자막은 progressive 캡션으로 인해 중복이 많음

SKILL.md

npx skills add https://github.com/aiskillstore/marketplace --skill youtube-transcript

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

[![Listed on Skillselion](https://skillselion.com/badge/skills/aiskillstore/marketplace/youtube-transcript.svg)](https://skillselion.com/skills/aiskillstore/marketplace/youtube-transcript)

Installs	1
repo stars	★ 404
Last updated	August 5, 2026
Repository	aiskillstore/marketplace ↗

What it does

Download YouTube subtitles with yt-dlp and turn them into clean transcripts or summaries.

Who is it for?

Extracting and cleaning YouTube subtitles into usable transcripts

Skip if: Videos with no captions where large Whisper transcription is undesirable

When should I use this skill?

You need a transcript or summary from a YouTube video

What you get

A clean, deduplicated transcript ready for summaries and chapter breakdowns

Clean transcript text
Optional summary and chapter breakdown

By the numbers

5-step priority order from yt-dlp check to Whisper fallback

Files

SKILL.mdMarkdownGitHub ↗

YouTube Transcript Downloader

유튜브 영상에서 자막을 추출하고 처리하는 스킬입니다.

Priority Order

1. yt-dlp 설치 확인
2. 사용 가능한 자막 목록 확인
3. 수동 자막 우선 시도
4. 자동 생성 자막 폴백
5. 최후 수단: Whisper 변환

Requirements

# macOS
brew install yt-dlp

# Linux
sudo apt install yt-dlp

# Universal
pip install yt-dlp

Workflow

Step 1: 자막 목록 확인

yt-dlp --list-subs "VIDEO_URL"

Step 2: 수동 자막 다운로드 (권장)

# 한국어 자막
yt-dlp --write-sub --sub-lang ko --skip-download "VIDEO_URL"

# 영어 자막
yt-dlp --write-sub --sub-lang en --skip-download "VIDEO_URL"

Step 3: 자동 생성 자막 (폴백)

yt-dlp --write-auto-sub --sub-lang ko --skip-download "VIDEO_URL"

Step 4: VTT → 텍스트 변환

# VTT 파일에서 타임스탬프 제거
sed '/^[0-9]/d; /^$/d; /-->/d' subtitle.ko.vtt > transcript.txt

Output Processing

중복 제거

자동 생성 자막은 progressive 캡션으로 인해 중복이 많음:

# 중복 라인 제거
seen = set()
unique_lines = []
for line in lines:
    if line not in seen:
        seen.add(line)
        unique_lines.append(line)

요약 생성

추출된 자막으로:

핵심 내용 요약
타임스탬프별 챕터 생성
주요 키워드 추출

Examples

Example 1: 강의 영상 요약

User: 이 유튜브 강의 요약해줘 - https://youtube.com/watch?v=xxx

Claude:
1. yt-dlp로 자막 다운로드
2. VTT → 텍스트 변환
3. 핵심 내용 요약 생성
4. 타임스탬프별 목차 제공

Example 2: 다국어 자막 추출

User: 이 영상의 영어/한국어 자막 둘 다 추출해줘

Claude:
1. --list-subs로 가용 언어 확인
2. 각 언어별 자막 다운로드
3. 정리된 텍스트 파일 제공

Error Handling

에러	원인	해결책
`yt-dlp not found`	미설치	brew/apt/pip 설치
`No subtitles available`	자막 없음	Whisper 사용 제안
`Invalid URL`	URL 오류	URL 형식 확인
`Video unavailable`	비공개/삭제	사용자에게 알림

Whisper Fallback

자막이 전혀 없는 경우 (사용자 확인 필요):

# 파일 크기 확인
yt-dlp --print filesize "VIDEO_URL"

# 사용자 승인 후 오디오 다운로드
yt-dlp -x --audio-format mp3 "VIDEO_URL"

# Whisper로 변환
whisper audio.mp3 --language ko --model base

주의: 대역폭/처리 시간 소요로 사용자 확인 필수

{
  "schema_version": "2.0",
  "meta": {
    "generated_at": "2026-01-17T04:22:45.937Z",
    "slug": "doyajin174-youtube-transcript",
    "source_url": "https://github.com/Doyajin174/myskills/tree/main/.public/skills/youtube-transcript",
    "source_ref": "main",
    "model": "claude",
    "analysis_version": "3.0.0",
    "source_type": "community",
    "content_hash": "4e8ed78ace5c662758c170f5dfd2e27951ef4b12d92d990c5e39bc4c28748add",
    "tree_hash": "b54778d963984d54726de44ed556c284659aa3c444c5db3d24aa1336843a979b"
  },
  "skill": {
    "name": "youtube-transcript",
    "description": "Download and process YouTube video transcripts using yt-dlp. Use this when extracting subtitles, creating summaries from videos, or processing video content.",
    "summary": "Download and process YouTube video transcripts using yt-dlp. Use this when extracting subtitles, cre...",
    "icon": "📹",
    "version": "1.0.0",
    "author": "Doyajin174",
    "license": "MIT",
    "category": "documentation",
    "tags": [
      "youtube",
      "transcript",
      "captions",
      "video",
      "yt-dlp"
    ],
    "supported_tools": [
      "claude",
      "codex",
      "claude-code"
    ],
    "risk_factors": [
      "network",
      "filesystem",
      "external_commands"
    ]
  },
  "security_audit": {
    "risk_level": "safe",
    "is_blocked": false,
    "safe_to_publish": true,
    "summary": "This skill consists solely of documentation describing yt-dlp usage patterns for downloading YouTube transcripts. All 38 static findings are false positives: cryptographic algorithm detections are random string false positives in JSON metadata; external command detections are documentation examples, not executable code; C2 keyword detection is a SHA256 hash being misidentified. The skill contains no executable code, no network operations beyond what yt-dlp performs, and no file access beyond standard Claude tool permissions.",
    "risk_factor_evidence": [
      {
        "factor": "network",
        "evidence": [
          {
            "file": "skill-report.json",
            "line_start": 6,
            "line_end": 6
          },
          {
            "file": "SKILL.md",
            "line_start": 94,
            "line_end": 94
          }
        ]
      },
      {
        "factor": "filesystem",
        "evidence": [
          {
            "file": "skill-report.json",
            "line_start": 6,
            "line_end": 6
          }
        ]
      },
      {
        "factor": "external_commands",
        "evidence": [
          {
            "file": "SKILL.md",
            "line_start": 17,
            "line_end": 23
          },
          {
            "file": "SKILL.md",
            "line_start": 23,
            "line_end": 27
          },
          {
            "file": "SKILL.md",
            "line_start": 27,
            "line_end": 36
          },
          {
            "file": "SKILL.md",
            "line_start": 36,
            "line_end": 42
          },
          {
            "file": "SKILL.md",
            "line_start": 42,
            "line_end": 44
          },
          {
            "file": "SKILL.md",
            "line_start": 44,
            "line_end": 48
          },
          {
            "file": "SKILL.md",
            "line_start": 48,
            "line_end": 54
          },
          {
            "file": "SKILL.md",
            "line_start": 54,
            "line_end": 58
          },
          {
            "file": "SKILL.md",
            "line_start": 58,
            "line_end": 60
          },
          {
            "file": "SKILL.md",
            "line_start": 60,
            "line_end": 64
          },
          {
            "file": "SKILL.md",
            "line_start": 64,
            "line_end": 67
          },
          {
            "file": "SKILL.md",
            "line_start": 67,
            "line_end": 74
          },
          {
            "file": "SKILL.md",
            "line_start": 74,
            "line_end": 82
          },
          {
            "file": "SKILL.md",
            "line_start": 82,
            "line_end": 93
          },
          {
            "file": "SKILL.md",
            "line_start": 93,
            "line_end": 101
          },
          {
            "file": "SKILL.md",
            "line_start": 101,
            "line_end": 104
          },
          {
            "file": "SKILL.md",
            "line_start": 104,
            "line_end": 111
          },
          {
            "file": "SKILL.md",
            "line_start": 111,
            "line_end": 117
          },
          {
            "file": "SKILL.md",
            "line_start": 117,
            "line_end": 118
          },
          {
            "file": "SKILL.md",
            "line_start": 118,
            "line_end": 119
          },
          {
            "file": "SKILL.md",
            "line_start": 119,
            "line_end": 120
          },
          {
            "file": "SKILL.md",
            "line_start": 120,
            "line_end": 126
          },
          {
            "file": "SKILL.md",
            "line_start": 126,
            "line_end": 135
          },
          {
            "file": "SKILL.md",
            "line_start": 32,
            "line_end": 32
          }
        ]
      }
    ],
    "critical_findings": [],
    "high_findings": [],
    "medium_findings": [],
    "low_findings": [],
    "dangerous_patterns": [],
    "files_scanned": 2,
    "total_lines": 313,
    "audit_model": "claude",
    "audited_at": "2026-01-17T04:22:45.937Z"
  },
  "content": {
    "user_title": "Extract YouTube video transcripts",
    "value_statement": "Manually created captions often have better quality than auto-generated ones. This skill provides a systematic approach to download YouTube subtitles using yt-dlp with fallback options for videos without captions.",
    "seo_keywords": [
      "YouTube transcript",
      "YouTube captions download",
      "yt-dlp",
      "Claude transcript extraction",
      "Codex video subtitles",
      "Claude Code YouTube",
      "extract subtitles",
      "video transcription",
      "subtitle downloader",
      "YouTube subtitles"
    ],
    "actual_capabilities": [
      "Download YouTube video transcripts in multiple languages",
      "Prioritize manually created subtitles over auto-generated ones",
      "Convert VTT subtitle files to plain text",
      "Handle videos without captions using Whisper fallback"
    ],
    "limitations": [
      "Requires yt-dlp to be installed on the system",
      "Cannot extract transcripts from private or deleted videos",
      "Auto-generated captions may contain errors"
    ],
    "use_cases": [
      {
        "target_user": "Content creators",
        "title": "Research and reference",
        "description": "Extract transcripts from educational videos for quotes and references"
      },
      {
        "target_user": "Developers",
        "title": "Video summarization",
        "description": "Download captions to create summaries and chapter outlines from long videos"
      },
      {
        "target_user": "Language learners",
        "title": "Multilingual content access",
        "description": "Get transcripts in native and target languages for study materials"
      }
    ],
    "prompt_templates": [
      {
        "title": "Basic transcript",
        "scenario": "Get video transcript",
        "prompt": "Extract the transcript from this YouTube video: [VIDEO_URL]"
      },
      {
        "title": "Multi-language",
        "scenario": "Get multiple languages",
        "prompt": "Extract both English and Korean subtitles from this video: [VIDEO_URL]"
      },
      {
        "title": "Clean output",
        "scenario": "Get plain text",
        "prompt": "Download the transcript and convert it to clean text without timestamps: [VIDEO_URL]"
      },
      {
        "title": "Full processing",
        "scenario": "Complete workflow",
        "prompt": "Download the transcript, remove duplicates, create a summary with key topics and timestamps: [VIDEO_URL]"
      }
    ],
    "output_examples": [
      {
        "input": "Extract the transcript from this YouTube lecture video about Python programming",
        "output": [
          "Video: Introduction to Python Functions",
          "Language: English (manual captions)",
          "Duration: 45 minutes",
          "Summary: This lecture covers defining functions, parameters, return values, and best practices",
          "Key topics: function syntax, default arguments, docstrings"
        ]
      },
      {
        "input": "Get both English and Korean subtitles from this K-pop documentary",
        "output": [
          "Available languages: English, Korean",
          "English captions saved to: video.en.vtt",
          "Korean captions saved to: video.ko.vtt",
          "Both files converted to plain text"
        ]
      }
    ],
    "best_practices": [
      "Prefer manual captions over auto-generated for better accuracy",
      "Verify video availability before attempting download",
      "Request user confirmation before using Whisper fallback for long videos"
    ],
    "anti_patterns": [
      "Attempting to download from private or unavailable videos",
      "Using auto-generated captions when manual ones are available",
      "Processing large videos without checking file size first"
    ],
    "faq": [
      {
        "question": "What languages are supported?",
        "answer": "YouTube supports many languages. Use --list-subs to see available options for each video."
      },
      {
        "question": "What if a video has no captions?",
        "answer": "Try Whisper fallback. This downloads the audio and uses OpenAI Whisper for transcription."
      },
      {
        "question": "Can I use this with Claude Code?",
        "answer": "Yes. Install yt-dlp first, then Claude can execute the transcript extraction workflow."
      },
      {
        "question": "Is my data safe?",
        "answer": "Yes. The skill only accesses YouTube public data. No personal information is collected or stored."
      },
      {
        "question": "Why do auto-captions have duplicates?",
        "answer": "Progressive captions on YouTube repeat content. The skill includes deduplication logic for cleaner output."
      },
      {
        "question": "How is this different from other transcript tools?",
        "answer": "This skill prioritizes manually created captions for accuracy and provides systematic fallback options."
      }
    ]
  },
  "file_structure": [
    {
      "name": "SKILL.md",
      "type": "file",
      "path": "SKILL.md",
      "lines": 138
    }
  ]
}