Skip to content

Video Skills for YouTube Shorts, Reels and TikTok

Vertical short-form video with word-by-word captions, kinetic titles and AI clips.

Skills for this

  • Code · video · free

    Animated Captions

    Word-by-word animated captions for talking videos. Transcribes and renders locally, free.

    Download · Free on GitHub

  • Code · video · free

    Kinetic Type Title

    Turn a line of text into an animated title clip. Renders locally, no API cost.

    Download · Free on GitHub

  • AI model · video · premium

    Product in Scene Ad

    One product photo in, a short ad clip with your product in a new scene out.

    Coming soon

  • Code · video · free

    Data-Chart Video

    Labels and numbers in, an animated bar or line chart video out. Free, renders locally.

    Download · Free on GitHub

Short-form video rewards speed and volume: a clear hook in the first seconds, readable text, and enough posts to learn what works. Editing each one by hand doesn’t scale. These skills turn the repeatable parts of a Short, Reel or TikTok into recipes your coding agent runs for you.

What you can make

  • Word-by-word captions. Animated Captions puts a caption block on your clip and highlights each word as it is spoken, the style viewers expect on short-form platforms. Pick where the block sits: bottom, middle or top.
  • Hook titles and end cards. Kinetic Type Title renders a line of moving text as its own clip: the hook at the start, or the call to action at the end.
  • Stats that move. Data-Chart Video turns your numbers into bars that grow and count up, in a vertical frame for Shorts and Reels.
  • B-roll from a photo. Product in Scene Ad turns one product photo into a short clip in the scene you describe, for the moments where you need something moving on screen.

Why code video for captions and titles

Captions and titles are text, and text must be exact. AI video models still garble letters, but a code video skill writes the frames as HTML and renders them with HyperFrames on your computer, so every word is spelled right and every caption lands on its beat. Change one word, render again, and the timing and colours stay the same. There is no API cost for a code render.

A workflow that scales

  1. Record or collect your footage at 9:16.
  2. Ask your agent to run Animated Captions on the clip; it transcribes the speech itself.
  3. Check the caption lines for misheard words before you render.
  4. Ask for a kinetic title for the hook and an end card with your handle.
  5. Post, look at which hooks keep people watching, and render new versions of the winners.

Format basics

Keep it vertical: YouTube treats square or vertical videos up to three minutes as Shorts, and a 9:16 frame fills the screen on every short-form app. Keep important text away from the very bottom and the right edge, where the platforms draw their buttons; that is why the caption position can be moved.

For ads in this format, see UGC ads.

What it costs

  • Code skills: free. They render on your own computer, so there is no API cost.

See pricing

Models used

Questions

What size should a Short be?
Vertical or square. YouTube counts videos with a square or vertical aspect ratio up to three minutes long as Shorts. Vertical 9:16 fills a phone screen on Shorts, Reels and TikTok.
Are these skills free?
Animated Captions and Kinetic Type Title are free code video skills that you download and render on your computer, with no API cost. Product in Scene Ad calls an AI video model, so its online runs will cost credits at the model's API price plus our margin.
Can the captions be made from my audio?
Yes. Animated Captions transcribes the speech on your computer with a free local speech model and times each word. You can also give it an SRT or VTT file you already have.
Which coding agents can use the skills?
Claude Code, Codex and Cursor. A skill is a folder with a SKILL.md file that teaches the agent the recipe.

Sources