2.5k
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
3 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill lh-video-gen- _meta.json278 B
- SKILL.md3.0 KB
Overview
This skill generates vertical short videos (9:16) directly from a Markdown script. It parses script sections into visual cards and speech, produces TTS audio for each segment, renders subtitle cards as images, and composites everything into a single MP4 using FFmpeg. The workflow is designed for fast, repeatable production of short-form social videos.
How this skill works
The tool parses a Markdown script that divides the video into segments using --- separators and expects fields for scene description, spoken text, and subtitle text. For each segment it generates TTS audio, creates a 9:16 subtitle/card image (or loads a provided image), then uses FFmpeg to merge the image with the audio into a clip. Finally it concatenates all clips into one vertical MP4 output.
When to use it
- Convert a scripted Markdown outline into a ready-to-share vertical short video.
- Automate production of narrated slide-style videos for social platforms.
- Batch-produce videos when you have consistent layout and pacing requirements.
- Quickly prototype short video content without manual editing tools.
- Combine with external TTS or image generation workflows for custom voices or visuals.
Best practices
- Write clear, short segments in the script so each card and TTS clip remains concise.
- Name pre-made images as slide_01.png, slide_02.png… to match script order when using --images-dir.
- Test TTS command templates locally before batch runs to confirm placeholders {text}, {output}, {voice}, {rate}.
- Keep subtitle text short and use explicit \n for line breaks to control card layout.
- Use --keep-temp during debugging to inspect intermediate images and audio files.
Example use cases
- A marketing short with a title card, three feature cards, and a CTA ending.
- Educational micro-lessons where each script segment covers a single concept.
- Repurposing blog posts into vertical videos by converting sections into script segments.
- Rapid A/B testing of voice and subtitle styles by swapping TTS commands or images.
- Archiving script-to-video outputs for batch backups or automated publishing pipelines.
FAQ
Yes. Provide a custom command with --tts-command using placeholders {text}, {output}, {voice}, and {rate}.
Do I have to use Chrome for generating images?
No. Chrome is used to screenshot HTML subtitle cards by default, but you can pass --images-dir with pre-rendered 9:16 images to skip Chrome.