Article Video Producer

Turn an article, essay, or Markdown note into a narrated video package: spoken script, AI voiceover, sidecar SRT subtitles, generated images, and a final MP4.

What It Does

• Rewrites long-form text into a more natural spoken script while preserving the original structure and style.
• Generates narration with Qwen3-TTS.
• Supports either cloning your own permitted voice or using a built-in model voice.
• Lets the AI adjust tone, pacing, and emphasis from the meaning of the article when using model voices.
• Creates external .srt subtitles, not burned-in captions.
• Plans documentary-style images, generates scene visuals, and renders the final video.
• Pauses for review after the spoken script and again before image generation.

Sample

Click the preview to watch the Gettysburg Address sample video:

• Download the sample SRT
• Source text: samples/gettysburg-address.md

GitHub README pages do not reliably render Release MP4 files as inline video players, so this README uses a clickable preview image.

How To Use

Install the skill by copying this folder into your agent skills directory:

mkdir -p ~/.agents/skills
cp -R article-video-producer ~/.agents/skills/article-video-producer

Then ask Codex in plain language:

Use the article-video-producer skill to turn my article into a video using my own voice.

Or:

Use article-video-producer to make samples/gettysburg-address.md into a narrated video. Use a model voice and let the AI choose the right emphasis.

For voice cloning, provide a clean 30-90 second reference recording and an exact transcript. Only clone voices you own or have explicit permission to use.

Outputs

Each run creates a fresh folder under:

<VaultRoot>/video/<article-stem>-video/

The main deliverables are:

video/final.mp4
video/final.srt

License

MIT. See LICENSE.