Text To Lottie logo

Text To Lottie

OrganizationPopular
diffusionstudio
text-to-lottie

Create, edit, or fix Lottie/Bodymovin JSON animations for the local Skia Skottie player. Use for text-to-Lottie, SVG/logo/type animation, loaders/icons, state feedback, UI microinteractions, lower thirds, diagrams, data/stat/chart animations, product promos, scene/camera motion, visual effects, scene edits, slots/controls, and Skottie debugging.

Overview

Publisherdiffusionstudio
Repositorylottie
Skill nametext-to-lottie
Stars
5.5K
Forks
290
Bundled files
22
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.

  • 22 bundled files

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

  • Open source

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

Installation

Install the Text To Lottie 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/diffusionstudio/lottie/tree/main/skills/text-to-lottie
  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/diffusionstudio/lottie.git /tmp/lottie
mkdir -p .claude/skills
cp -r /tmp/lottie/skills/text-to-lottie .claude/skills/text-to-lottie
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Text To Lottie 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 Text To Lottie 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 Text To Lottie 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.

Text To Lottie

Author production-ready Lottie JSON for the official local Skia Skottie player. The deliverable is a renderable scene in the player, not isolated JSON.

Operating Model

  • Use the official player project and verify in Skia Skottie. Do not hand-roll a custom viewer or switch renderers for verification.
  • Keep the skill portable across Agent Skills clients. Avoid host-specific commands, command modes, or orchestration conventions in skill instructions.
  • Prefer fewer questions and stronger defaults. Ask only when a decision changes the output materially, such as transparent vs full-frame background, brand constraints, target format, or supplied source assets.
  • Prioritize clean, intentional, professional motion over merely satisfying the literal prompt.

Reference Loading

This SKILL.md is the thin control plane. Load only one-level references that match the task. Do not open the whole reference library.

Always read references/player-contract.md before creating, editing, fixing, or verifying a scene. If a routed reference is unavailable, continue using the inline rules in this file.

User intentReferences to read when present
Any new/edit/fix Lottie scenereferences/player-contract.md
JSON structure, keyframes, slots, shapes, assetsreferences/lottie-spec-map.md
Logo animationreferences/recipe-logo.md, references/motion-taste.md, references/design-taste.md
Typography, title, quote, text revealreferences/recipe-typography.md, references/design-taste.md, references/motion-taste.md
Lower third, name tag, caption bar, overlayreferences/recipe-lower-thirds.md, references/design-taste.md, references/motion-taste.md
Loader, icon, spinner, badge animationreferences/recipe-loaders-icons.md, references/motion-taste.md
Success, error, warning, completion, empty statereferences/recipe-loaders-icons.md, references/design-taste.md, references/motion-taste.md
UI microinteractionreferences/recipe-ui-microinteractions.md, references/design-taste.md, references/motion-taste.md
Generic "animate this SVG" or SVG-to-Lottiereferences/recipe-svg-animation.md, references/svg-compatibility.md, references/motion-taste.md
Camera follow, pan, zoom, parallax, scene motionreferences/recipe-camera-scene-motion.md, references/design-taste.md, references/motion-taste.md
Diagram, technical line animation, callout, flow tracereferences/recipe-diagram-technical.md, references/design-taste.md, references/motion-taste.md
Data, stats, KPIs, charts, metrics, dashboard figuresreferences/recipe-data-stats.md, references/design-taste.md, references/motion-taste.md
Product launch, feature announcement, social promoreferences/recipe-product-promo.md, references/design-taste.md, references/motion-taste.md
Long text, multiple ideas, list/features/steps, timeline, before/after, problem/solution, quote+proof, recap/story, product walkthrough, multi-language variations, chapters, multi-beat sequence, episode, jump/hard cuts, transition grammarreferences/chapterization-transition-grammar.md, references/motion-taste.md
Glow, glass, metal, gradient, fill, bubble/burst effectsreferences/recipe-visual-effects.md, references/design-taste.md, references/motion-taste.md
SVG input inside logo/icon/UI/lower-third workThe task recipe plus references/svg-compatibility.md
Starter brief or reusable project directionreferences/recipe-starter-projects.md, references/design-taste.md, references/motion-taste.md
Any "premium", "clean", "minimal", "modern", "sleek", or "polished" qualifierreferences/design-taste.md (restraint defaults), plus the routed recipe

For mixed prompts, choose one primary recipe from the main deliverable, then add secondary references for source format or visual treatment. Examples: SVG logo with glow uses logo as primary plus SVG compatibility and visual effects; product launch with pan/zoom uses product promo plus camera scene motion; technical SVG diagram tracing uses diagram/technical plus SVG compatibility; kinetic title with glass sweep uses typography plus visual effects.

Workflow

  1. Route the task using the table above. Read only the relevant references that exist.
  2. Locate the official player project and resolve the target scene using the target precedence below. Before editing, verify the resolved path is public/projects/<project>/<scene-N>/lottie.json; re-read that current file immediately before overwriting because the UI can write slot edits back to source.
  3. Decide the background policy before authoring.
  4. Write or update public/projects/<project>/<scene-N>/lottie.json and, when useful, controls.json.
  5. Validate JSON, run or reuse the dev server, inspect exact frames with ?frame=N, and fix render/design/motion issues before finishing.

Inline Rules

Design Defaults (always apply)

These few defaults are non-negotiable and apply to every designed scene. Load references/design-taste.md for the full reasoning, especially for any "premium", "clean", "minimal", "modern", or "card" prompt.

  • Premium means subtract, not add. When a prompt says premium, clean, minimal, modern, sleek, or sophisticated, default to restraint: remove chrome before adding it. Premium is carried by scale, weight, brightness, spacing, and timing, never by cards, borders, dividers, shadows, glow, or stacked tints.
  • Chrome/container budget is 0 by default. Do not add a framing card, container, border, or divider unless it does a job that whitespace and alignment cannot. Separate grids and columns with negative space and alignment first, a single hairline second, and a filled or bordered card last and only when explicitly warranted.
  • One surface tone. Use one background tone. Do not stack two near-black or two near-white tints to fake a "surface"; it reads muddy. If a card surface must differ from the background, make it one deliberate step with a clear purpose.
  • One divider treatment, one color. If dividers are truly needed, use one weight and one color for all of them, title rule and column rules included. Never use slightly different colors or weights per divider.

Scene Rules

  • Scene files live at public/projects/<project>/<scene-N>/lottie.json.
  • Resolve target scenes by authority: an explicit file path wins; a browser URL route like /<project>/<scene> wins next; an already-known project/scene for the task wins next; otherwise create a new safe scene. Use /__context only for discovery, verification, or playback state unless the task explicitly says to edit what is currently on screen and no more specific target exists. Do not let /__context.live override a known file path, URL, or project/scene. Overwrite main-project/scene-1 only if it is still the untouched placeholder; otherwise create a new scene.
  • Full-frame standalone compositions should include a visible background layer with a bgColor slot and controls.json entry.
  • Transparent-by-default outputs include logos, icons, loaders, overlays, lower thirds, and SVG-derived assets unless the user asks for a background.
  • Include top-level v, fr, ip, op, w, h, nm, assets, and layers. Treat op as exclusive.
  • Use purposeful easing and staging. Avoid defaulting to linear motion. Derive easing from the behavior-based anchors in motion-taste.md (chosen by motion behavior, focal element strongest); do not fall back to one uniform ease for every layer.
  • Use slots for important editable values and add controls.json labels/ranges when they improve the properties panel.
  • For SVG input, preserve the viewBox, normalize styling, watch fill rules and intersections, and verify the result in Skottie.
  • Native Lottie text/text slots (ty:5) render in this player when the scene ships its font: drop a .ttf/.otf/.ttc next to lottie.json, declare it in fonts.list with fFamily matching the font's embedded family name, and reference it from text documents. The loader passes every scene font to Skottie. Prefer native text; reach for vector/shape text only for deliberate path effects (stroke-on, glyph morphs, handwriting). See the player-contract "Native Text" reference.

Verification

Before finishing:

  1. Confirm the intended target file path is public/projects/<project>/<scene-N>/lottie.json.

  2. Validate JSON:

    bash
    node -e "JSON.parse(require('fs').readFileSync('public/projects/<project>/<scene-N>/lottie.json','utf8'))"
  3. Confirm the official player is running and the scene appears in GET /__context.

  4. Inspect pinned frames in the browser. For new scenes, check frame 0, midpoint, and op - 1.

  5. Confirm the background policy matches the use case.

  6. Check for blank canvas, missing assets, unstyled shapes, wrong layer order, bad easing, awkward timing, cropped content, text overflow, and visible SVG artifacts.

  7. Finish only when the animation renders cleanly and feels intentional.

Maintenance Evals

Do not read eval files during normal animation work. Use them only when testing or changing this skill:

  • evals/trigger-prompts.json
  • evals/routing-prompts.json
  • evals/reference-loading-prompts.json
  • evals/output-rubric.md

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 Text To Lottie AI skill do?

Create, edit, or fix Lottie/Bodymovin JSON animations for the local Skia Skottie player. Use for text-to-Lottie, SVG/logo/type animation, loaders/icons, state feedback, UI microinteractions, lower thirds, diagrams, data/stat/chart animations, product promos, scene/camera motion, visual effects, scene edits, slots/controls, and Skottie debugging.

Why use Text To Lottie on TypingMind?

Because you install it once and use it with any model. Text To Lottie 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 Text To Lottie in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/diffusionstudio/lottie/tree/main/skills/text-to-lottie. 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 Text To Lottie?

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 Text To Lottie?

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

Is the Text To Lottie AI skill free?

Yes. It is published on GitHub by diffusionstudio 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 👇