
Shorts to Douyin Video Workflow Pack
Preserve the core story. Localize the presentation.
This downloadable skill provides production instructions for a compatible AI agent. It is not a standalone model or a one-click video generator. Analyze a YouTube Short, lock its setup, essential characters and props, key actions, sequence, contrast, and ending, then recreate it using newly generated or self-shot footage—not downloaded copies of the original or an unrelated replacement story.
What the workflow covers
- Source verification, direct video analysis, core-story planning, and deduplication by both source and story.
- Fictional adult Chinese characters where the story calls for them, with consistent appearance and voice across shots; no celebrity face swaps or real-person voice cloning.
- Storyboarding and production defaults: 9:16, 720 × 1280, usually 15–30 seconds, with the core and ending taking priority over padding or arbitrary duration.
- Natural Mainland Chinese Mandarin, large Simplified Chinese captions timed to the actual finished footage, and readability and obstruction checks.
- Music with verified rights for embedding in the finished video, or original Douyin-style instrumental music. Suitable ASMR or reaction videos may use no background music.
- A brief closing call to action after the complete ending: “点赞、订阅、分享” on screen, with a Mandarin reminder. It does not replace the story ending.
- Content and audio checks, MP4 finishing, asset organization, and delivery guidance with a Simplified Chinese title and short caption.
Your download includes 6 files
SKILL.md— main workflow, quality checks, and operating boundaries.references/core-fidelity.md— story criteria, correct/incorrect examples, and historical notes.references/production-standard.md— visual, Mandarin, caption, music, closing-CTA, and finishing standards.scripts/finish_clip.py— Python/FFmpeg finishing script for completed raw footage, approved caption timings, and optional music.templates/clip-plan.json— sources, core-story planning, characters, shots, and checks.templates/captions.sample.json— sample post-production caption JSON.
Language and compatibility
The workflow files are written in Traditional Chinese. Use a multilingual agent capable of understanding those instructions. An English product listing does not mean an English translation of the workflow is included. The intended video output is Mandarin speech and Simplified Chinese captions, not English-language videos.
You need a compatible AI agent with source verification, video analysis, image/video generation, TTS, and music tools; Python 3, FFmpeg, and ffprobe; ASS subtitle rendering support; and a CJK font that renders Simplified Chinese. Supply your own tool accounts, credits, and permissions. Features, quality, and costs depend on your chosen tools. Verify compatibility before purchase.
Included examples and historical notes are not your personal production history or a live list of trending videos.
Quick start
After downloading, read SKILL.md and both guides, then run the finishing script with --help. Provide a Shorts link you have the right to use and ask your agent:
Analyze this YouTube Short, lock the core event, essential actions, and ending, then follow the skill to recreate it as a 9:16 video with Mandarin and large Simplified Chinese captions. Deliver an MP4, title, and short caption.
Completed videos require working external tools, available credits, and any necessary confirmations. Missing tools should be disclosed rather than presenting a storyboard or image as a finished video.
Price and important limits
US$5 — one-time purchase of the downloadable skill files. Model usage, video/image generation, TTS, music, hosting, other third-party service fees, and done-for-you videos are not included.
The finishing script does not research sources, generate footage, translate captions, automatically align timings, or create the spoken CTA. Prepare completed footage and approved caption timings first. The skill does not start daily runs or publish by itself, and grants no account access or publishing permissions. Scheduling and publishing require separate setup and any required confirmations. No guarantees of video volume, generation success, views, revenue, or platform approval.
Localization does not automatically grant rights to footage, distinctive story expression, likenesses, or music. Verify lawful use. No copyright exemption or unauthorized songs are included. Preserve existing provider/platform labels and source information, and make applicable AI-content disclosures when publishing. Watermark-free or AI-label-free output is not promised.
For creators and AI agent users who already have media tools and want a consistent Chinese short-video localization workflow. Support: [email protected].
واجهات API الخارجية
AI Agent & Media Tools / AI 影音工具


