Producing shorts
Skill dpyeye-commits/my-arsenal-marketplace/skills/producing-shorts
Produces 5+1 stage YouTube Shorts (search, 10 title variants, 60s script with 5 hook rules, TTS, 15-cut image prompts, 3 thumbnail variants). Use when the user requests 쇼츠, 숏폼 영상 기획, 유튜브 쇼츠, or 60초 대본.From its SKILL.md
npx -y skills add dpyeye-commits/my-arsenal-marketplace --skill producing-shortsAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.6 KB, 649 tokens by cl100k_base, as published. Nobody here has run it
Shorts Pipeline (5+1 단계)
0단계 사전 확인 (환각 방지)
- 채널 (영상치유lab / 경제학 / 기타) — 톤·페르소나 분기
- 주제 키워드 1~2개
- 목표 길이 (기본 60초)
1단계 — YouTube 검색
- YouTube Data API v3 (
search.list+videos.list) - 상위 10개 제목·설명·조회수·게시일
2단계 — 제목 10안 (후킹 5규칙 강제)
- 3초 룰 (첫 3초 안에 후킹)
- 오픈 루프 3개
- 숫자 1개 이상
- 2인칭 ("당신")
- 부정 → 긍정 플립
3단계 — 60초 대본 (3막 압축)
- 1막(0~10초): 후킹 + 갈등 제시
- 2막(10~45초): 우화·예시·핵심 메시지
- 3막(45~60초): 학술 인용 + CTA
4단계 — TTS
- edge_tts (msedge-tts npm) 무료
- Google Cloud TTS (Neural2)
- GPT-SoVITS (1분 클로닝)
5단계 — 이미지 프롬프트 15컷
- 60초 ÷ 4초 = 15컷
- 각 컷: 장면·캐릭터 시드·카메라·조명·스타일
- 스타일 락 (영상치유lab=Studio Ghibli 애니 / 경제학=클레이)
6단계 — 썸네일 3안
- A: 후킹 텍스트 大
- B: 캐릭터 정면
- C: 숫자·통계 강조
추측 금지
- 데이터에 없는 통계·인용 절대 생성 X
- peer-review 인용은 CrossRef DOI 검증 후만
무기고 페어
- skills/youtube/shorts-pipeline-app.md
- skills/youtube/youtube-shorts-strategy.md
- skills/cinema/screenplay-writing.md
What ships with it: 2 files
3.9 KB alongside SKILL.md, 1 of them executable
evals/
- evals.json1.5 KB
- run.mjsruns2.4 KB