smart-short-video

Automating short video creation involves slicing original footage into segments, generating AI images, transcribing audio, and mixing scenes before final rendering. This Skill coordinates these steps end-to-end to produce a 9:16 remix.

Remotion is used to render the final 9:16 short video. The Skill coordinates slicing, transcription, AI-image generation, scene mixing, and final Remotion rendering with guided prompts and safety checks.

Audio is transcribed using Whisper to produce concise copy for captions. This transcription step is integrated into the end-to-end short-video creation workflow.

Scene mixing and pacing utilize a Fisher-Yates shuffle and flexible mix modes to create engaging sequences. You can set target durations of 30, 60, 90, or 120 seconds.

The process supports multiple AI image services and randomization to accompany scenes. It integrates these generated images with original video segments during the scene mixing phase.

Automatic video slicing splits source video into fixed 3-second segments for modular editing. This rigid segmentation structure may not perfectly align with specific scene transitions or audio cues.