Agent skill · creative production · guanyang
video-tutorial-maker
Create scripted tutorial videos with narration, TTS, aligned subtitles, MP4 rendering, and platform variants. Use when Codex needs to create or revise tutorial, explainer, product walkthrough, course, demo, or short-form videos; generate 16:9 landscape output by default unless the user explicitly asks for 9:16, 1:1, or another format.
Why this skill is useful
Adds specific commands and workflows for generating tutorial videos that the AI wouldn't reliably generate on its own.
What it needs
Requires edge-tts, ffmpeg installed locally. About 3k tokens when loaded. Last updated 2026-07-16. 28 stars on the source repository.
What this skill does
Video Tutorial Maker Use this skill to turn a topic, outline, script, or existing video into a finished tutorial video package. Default assumption: produce a 16:9 landscape MP4 first. Create 9:16 or other platform variants only when requested or clearly useful. Workflow 1. Define the deliverable. Confirm the topic, target audience, tone, duration, language, and required output path when they are unclear. If the user does not specify aspect ratio, use 1920x1080 landscape video. For coding or software tutorials, prefer a clear slide/screen sequence over dense narration. 2. Write the video package. Create a concise script with one narration block per scene or slide. For each scene, define title, subtitle, bullets or screen action, visual style, and narration. Keep on-screen text shorter than narration. Do not force full captions into the visual design. 3. Generate speech. Prefer edge-tts when available because it is free and more natural than basic system voices. For Chinese, start with zh-CN-XiaoxiaoNeural; adjust voice or rate only if the user asks or the result is clearly poor. Generate audio per scene, not one large narration file. Per-scene audio makes transition sync and subtitle offsetting deterministic. 4. Build subtitles from the same timeline as audio. Use the subtitle output from TTS when available. Parse both 00:00:01.234 and 00:00:01,234 timestamp formats. Offset each scene subtitle by the cumulative scene duration. Verify the final subtitle file contains real timestamps and no NaN. 5. Render video. Render each scene to a fixed duration equal to its audio duration plus any intended pause. If a scene has a trailing pause, pad that scene's audio with silence to the exact scene video duration before concatenating. Concatenate padded scene audio, not raw scene audio. Otherwise voice will drift earlier after each transition. Mux video, audio, and soft subtitles into MP4 when supported; otherwise output the subtitle file as a sidecar. 6. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills guanyang/video-tutorial-maker