Animate logo

Animate

Organization
with-pebbly
animate

Build an idle, walk, run or attack cycle from key poses, set timing that reads as motion rather than a metronome, tag it, and review it as a filmstrip. Use when a rigged sprite needs to move.

Overview

Publisherwith-pebbly
Repositoryaseprite-ai-artist
Skill nameanimate
Stars
456
Forks
28
Bundled files
Instructions only
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by with-pebbly 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/with-pebbly/aseprite-ai-artist/tree/main/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/with-pebbly/aseprite-ai-artist.git /tmp/aseprite-ai-artist
mkdir -p .claude/skills
cp -r /tmp/aseprite-ai-artist/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 a cycle

Before you start

The sprite must be rigged onto named layers — see aseprite:rig. Animating a flattened sprite means redrawing every frame by hand.

Procedure

0. With a storyboard

If aseprite:concept delivered a storyboard, its panels are the key poses — the plan in step 1 is written from them, not invented. Import it once the frames exist: frame op="add" count=N-1, then reference op="import" path=… grid={columns: N, rows: 1} puts panel i on frame i. Every frame then has its own reference, and look op="compare" frame=i checks that frame against its panel — do that for each key in step 4. The PixelSpec's per-frame notes and deviations still decide what survives at this size: a storyboard can suggest a 4px jump the timing cannot carry.

1. Plan the key poses

Write them out before touching a frame. For a walk:

  1. Contact — front foot lands, back foot pushing. Widest stance.
  2. Down — weight on the front leg, body lowest.
  3. Pass — legs together, body highest. The frame everyone forgets, and its absence is why a walk looks like sliding.
  4. Up — pushing off, rising.

Then mirror for the other leg: 8 frames. A 4-frame walk uses contact and pass per leg.

Idle: 2–4 frames, a breath — chest rises a pixel, shoulders follow one frame later. Attack: anticipation, strike, hold, recover.

2. Create the frames

frame op="add" count=7

Already done if a storyboard was imported in step 0 — the import needs them.

3. Block the key poses

Move limbs by moving cels, not by redrawing:

cel op="copy" layer="arm-near" frame=1 toFrame=3
cel op="move" layer="arm-near" frame=3 dx=2 dy=-1

Redraw only where a part genuinely changes shape — a foreshortened arm, a bending knee. Do that redraw as a grid: look op="ascii" layer=… frame=… rulers=false region=… on the part in the previous key, edit the rows into the new shape, and draw kind grid them onto the new frame at the same origin. Copying the rows first keeps the part's volume from drifting, because every pixel you did not change stays exactly where it was.

4. Look at the keys before adding in-betweens

look op="filmstrip"

A vision model reads only the first frame of a GIF, so the filmstrip is the only way to actually see the motion. Check volume and height consistency here, while there are four frames to fix rather than eight. With a storyboard, look op="compare" frame=… on each key as well: the pose should match its panel before any in-between is drawn from it.

5. Add in-betweens

Only once the keys read. Spacing is your easing: poses close together read slow, far apart read fast.

look op="onion" frame=3

Use onion skinning to judge an in-between against its neighbours without stepping through frames one at a time — it composites the target at full opacity over ghosted neighbours, oldest-first.

For secondary motion — a bob, a breath, a float — cel op="tween" or cel op="oscillate" fill in the in-between cels for you instead of keying each one by hand:

cel op="oscillate" layer="hair-back" fromFrame=1 toFrame=8 amplitudeY=1 period=8

Caveat: easing rounds each frame's offset to the nearest pixel, and on small amplitudes (1–2px) that rounding can make two adjacent frames land on the same value and then jump — a stutter, not a stumble. Prefer hand keys for primary motion (the walk, the strike); reach for tween/oscillate for motion nobody is meant to consciously notice.

Something that grows or shrinks on screen — a wagon coming down a road, a bird flying away — is drawn at three or four sizes by hand (far, middle, near), each a complete little sprite with all its parts, and every frame uses the nearest one. Do not compute it from a formula and re-rasterise it at each size: below about 12 px a scaled shape loses or merges its parts (the wheels vanish, the cover becomes a blob), and it changes shape from frame to frame. Before you finish, look op="preview" frame=<n> on the frames where it is largest and smallest.

6. Set timing

frame op="set_duration" durations=[150,100,100,150,150,100,100,150]

Hold contact poses longer than pass poses. Uniform timing reads mechanical no matter how good the drawings are. Typical ranges are in rules://05-animation.

7. Tag it

tag op="create" name="walk" from=1 to=8 direction="forward"

Untagged frames are unusable by a game engine. validate treats a multi-frame untagged sprite as an error.

8. Review

look op="filmstrip" again, then run through the animation checks in rules://07-review-checklist: consistent volume, consistent height, clean loop, anticipation before strong actions, light that does not move. Run validate with an expect describing the cycle — which layers live on which frames, which parts must never overlap — so the mechanical check knows what "right" means here.

Related

rules://05-animation, rules://06-layers-and-rigging, rules://40-timing-and-spacing. For the cycle itself: rules://41-idle-and-breathing, rules://42-walk-and-run, rules://43-jump-fall-land, rules://44-attacks-and-impacts, rules://45-secondary-motion, rules://46-subpixel-animation, rules://47-top-down-animation; animals: rules://51-animal-gaits, rules://52-birds-and-flight. Their frame tables and key-pose templates come before your own invention.

Frequently asked questions

What does the Animate AI skill do?

Build an idle, walk, run or attack cycle from key poses, set timing that reads as motion rather than a metronome, tag it, and review it as a filmstrip. Use when a rigged sprite needs to move.

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/with-pebbly/aseprite-ai-artist/tree/main/skills/animate. TypingMind reads its SKILL.md 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 with-pebbly 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 👇