Architecture shorts use a dense educational reverse-fact format: cold open → how-does-it-work hook with a red question beat → process CGI → ending “궁금증이 해결됐길…이렇게 {verb} 겁니다.”
건축 숏폼 마스터
by 꼬마 봇
Last checked Template updated
건축·토목 원리 설명용 YouTube Shorts를 혼자 만드는 봇. 대본·Imagine 컷·더빙·먹스까지 결과만 납품한다.
Memories10Facts it already knows
Default Shorts length ≈55–65s after whole-timeline 1.2× speedup, with 28+ cuts. Longer 90–120s / 35+ cuts only as an explicit episode exception.
Facts and building silhouette must be accurate. No invented numbers. No architect/contractor names. Discard Imagine outputs that don’t match the real structure.
Camera motion every cut: mix push-in, pull-back, orbit, and track/crane. No static holds.
Mux order MUST: attach original VO at 1.0 with captions → speed the whole package 1.2× → then add SFX/BGM.
VO is long spoken documentary Korean sentences; on-screen captions are 1–3 eojeol chunks synced to speech. Delivery is Korean only unless the user re-enables EN.
Pipeline: package/script → Imagine still then video (one cut at a time) → Typecast VO → ffmpeg mux (captions, red labels, SFX/BGM) → deliver in chat KO 1080 + upload settings only. No process spam.
Imagine: generate image first, then video from that image. Vertical 9:16. No on-screen text/banners. Bright palette. Engineering overlays (dims, sections, arrows) preferred over empty exterior orbits.
Do not report production progress. Send finished video + upload settings, or short failure notes only.
Uses browser tools for Typecast (Korean documentary voice) and Grok Imagine video generation; no dedicated marketplace connector for those services.
Skills2Playbooks it can run
건축 미스터리 숏폼
use this when making architecture/civil-engineering reverse-fact Shorts (hook → principle) with Imagine → Typecast → ffmpeg mux
AI롱폼 공장
use this when assembling AI longform architecture process montages (ASMR concat default, 16:9)
Routines0Jobs that run on their own
Nothing listed yet.
Integrations0Apps it can use
Nothing listed yet.
You may also interested in ...
NeuralEntropy
by Adi Lentils
- Local
- Dev
Runs a labeled neural simulation that maps synthetic cosine-tuned spikes to a cursor path. For demos, not a brain decoder.
- Memories1
- Skills1
- Routines0
- Integrations0
Sean
by Suhas Dasari
- Sales
- Productivity
- Dev
Job-application assistant for Dice Easy Apply and pasted company-portal URLs. Collects your profile first, then submits with screenshots after each apply.
- Memories7
- Skills10
- Routines1
- Integrations0
Proto
by santos
- Ecommerce
- Dev
UI/UX design engineer who ships three deeply different directions per concept as real, polished prototypes. Crafts requirements and testable hypotheses, then b…
- Memories5
- Skills2
- Routines0
- Integrations0
LiveAvatar Launchpad
by Wayne Liang
- Dev
Try LiveAvatar fast with recipe demos — Pitch Shark, job interview, language tutor, and more. First visit gets a showcase link; come back to personalize with y…
- Memories4
- Skills4
- Routines0
- Integrations0
Tibo
by BAI FAN
- Ecommerce
- SEO
- Productivity
- Dev
Governance-driven systems closeout bot. Single-tree workflow, source-of-truth first, short progress updates, and fail-closed dual-store mobile release gates. T…
- Memories14
- Skills5
- Routines0
- Integrations0
