Animate logo

Animate

Organization
imgly
animate

Use when an existing CoDesign design should move — "animate this", "make it move", "turn this post/poster/carousel into a video", "motion version", "make a loop of this". Turns a still single- or multi-page design into one animated page and an mp4 without redesigning it, steered by one intake round and plain-language follow-ups ("slower", "logo last", "no bounce"). Not for videos from a repo or website (launch-video) or for new designs (create).

Overview

Publisherimgly
Repositorycodesign
Skill nameanimate
Stars
43
Forks
1
Bundled files
3
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by imgly on GitHub. Read the source before you install it.

Installation

Install the Animate AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/imgly/codesign/tree/main/plugins/codesign/skills/animate
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/imgly/codesign.git /tmp/codesign
mkdir -p .claude/skills
cp -r /tmp/codesign/plugins/codesign/skills/animate .claude/skills/animate
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Animate in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Animate on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Animate is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

animate

Give a finished still design motion. The layout never changes: every animation starts from the design and settles on it, so the settled frame IS the still. Builds on, without restating, the handbook loop, ../handbook/video.md, ../handbook/intake.md and the judge gate. Needs the native engine for mp4; if export refuses because the server runs on WASM, stop and report the reason.

1 — Read the design

Work from the design as it stands, and keep the still: save it (export({ format: 'imgly' })) and export a png of each page as the reference frames §5 compares against. Then animate a copy: import({ source: { uri } }) with the saved uri starts a new design from it, and the still stays as it was. Run reference/schedule.md's inventory.js block as one read-only edit with render: false: per page its id, every group, and every leaf with id, type, name, global bounding box, fill kind, enclosing group, and words and font size for text blocks. Never write your own inventory: reading text properties on an image throws. preview the pages.

Give each layer a key (s<page>/<short-name>) and a role:

rolewhat
backgroundbottom-most full-bleed layer (image, colour, gradient)
overlayany other full-bleed layer: a scrim or gradient over the background
imagepictures, illustration layers, product shots
headlinelarge display text
bodysmall text: overlines, captions, sublines, footers
decorationrules, pills, icons, shapes
logo · ctabrand mark · call to action

Record words and fontSize for text. List every group under its page's containers — a group left out hides its children after 5 s — and put every id (each page as s<n>, each layer, each group) under ids. For multi-page designs also a sig for chrome that repeats on every page — <type>|<text or fill>|<x,y,w,h rounded to 1 px> — so it is kept once.

2 — Intake (one round)

Ask only what the prompt leaves open, in your host's question UI, each with Decide automatically:

parameterchipsdefault
stylecalm · energetic · playful · cinematicfrom the design's tone
lengthshort ~6 s · medium ~10 s · long ~15 s · fit to contentfit to content
endinghold the final frame · loop · outrohold
herothe 2–3 strongest candidates by namelargest headline, else largest image
pages (multi-page only)all in order · pick pagesall in order

A loop is seamless only on a single page: a multi-page loop ends on the last page's background and cuts back to the first page, so say that when offering it there. Pages left out of a pick are dropped from the animated design; the saved still keeps them.

Echo the resolved inputs in one to three lines, marking what was inferred or defaulted, then build.

3 — Schedule

Run reference/schedule.md's schedule.js block as a read-only edit (render: false), its first line const INPUT = <input>, CHANGE = null; — the inventory in the shape documented there, with this session's block ids under ids. It returns the plan. Never set timings by hand — the block owns ordering, the settle rule (every entrance done by 60 % of its scene), reading time, endings, multi-page scene windows and which options each animation type accepts. If the plan says stretched: true, tell the user the video is longer than asked and why.

4 — Build — ONE edit

Code: const PLAN = <the plan>; followed by the apply.js block, run on the design you animate (§1), never on the saved still. Do not fix anything else in that edit. The server may list pre-existing lint findings on the source design (line height, tracking): they are the still's, not yours — never change text or layout properties while animating; mention them at delivery.

5 — Verify

  • Compute each scene's settle point from plan.apply: the latest timeOffset + in.duration over the scene's entries — every key starting s<n>/, including the background and shared chrome, which have no parent — plus the scene's content group timeOffset for entries with a parent (none on a single page). Do not use a fixed fraction of the length: loop and outro exits can start earlier.
  • preview the page with time at each settle point: each must look exactly like the still (page by page for multi-page).
  • Run the judge loop on those frames and record the scorecard.
  • preview the page with time at the middle of every transition and at the last frame (duration − 0.1). With ending hold, compare the last frame with the reference png from §1 — they must match. With loop or outro the last frame is only the background (everything has exited): compare the settle-point frame instead.
  • export({ format: 'mp4', blockId: page }) once those frames pass, and preview the last frame again right after it: black means the export went black (the judge skill's export check).

6 — Deliver

The mp4, a poster (png of the settled frame — park the playhead with setProps(page, { playback: { time } })), and the .imgly. Offer the follow-ups: "slower", " last", "no bounce", "more energy", "loop it", "add music" (the models skill).

Iterate

Read the page's codesign/motion metadata ({ v, input }) in a read-only edit, then run the schedule.js block again as a read-only edit with const INPUT = <that metadata>, CHANGE = <change>;:

user sayschange
slower / faster{"kind":"slower"} / {"kind":"faster"}
X first / X last{"kind":"first","key":"s1/logo"} / {"kind":"last",…}
no bounce{"kind":"noBounce"}
more energy / calmer{"kind":"energy"} / {"kind":"calmer"}
don't animate X{"kind":"static","key":"…"}
loop it / hold the end / outro{"kind":"ending","to":"loop"} …

Apply the result with the same apply.js edit on the animated design (the still has no motion-beat tags) — it finds blocks by their codesign/motion-beat tag, so no ids are needed, and re-applying replaces animations and transitions instead of stacking them. Block ids are session-scoped: a saved plan never carries them, so never paste ids stored in an earlier session. Then verify again. A request outside the table: edit that beat's entry in the plan JSON, re-run schedule.js on it, apply.

References: reference/styles.md (what each style does and why), reference/multipage.md (how pages become scenes).

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Animate AI skill do?

Use when an existing CoDesign design should move — "animate this", "make it move", "turn this post/poster/carousel into a video", "motion version", "make a loop of this". Turns a still single- or multi-page design into one animated page and an mp4 without redesigning it, steered by one intake round and plain-language follow-ups ("slower", "logo last", "no bounce"). Not for videos from a repo or website (launch-video) or for new designs (create).

Why use Animate on TypingMind?

Because you install it once and use it with any model. Animate is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Animate in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/imgly/codesign/tree/main/plugins/codesign/skills/animate. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Animate?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Animate?

As many as you like. As long as a model supports skills, you can use Animate with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Animate AI skill free?

Yes. It is published on GitHub by imgly under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇