An agent-oriented skill for turning an SVG into a hand-drawn animation deliverable.
This repository is intentionally small. It contains only the skill definition, reusable assets, and one example demo.
Default deliverables:
preview.htmlplayer.js
Default behavior:
- draw SVG paths first
- reveal fills afterward
- preserve source colors as closely as possible
.
├── README.md
├── .gitignore
├── examples
│ └── svg-player-demo
│ ├── 975549_OGFB9T1.svg
│ └── svg-player-demo.html
└── skills
└── svg-hand-drawn-preview
├── SKILL.md
├── assets
│ ├── player.js
│ └── preview-template.html
└── references
├── html-preview-guidelines.md
└── output-contract.md
- Skill entry:
skills/svg-hand-drawn-preview/SKILL.md - Reusable player asset:
skills/svg-hand-drawn-preview/assets/player.js - Preview template:
skills/svg-hand-drawn-preview/assets/preview-template.html - Output contract:
skills/svg-hand-drawn-preview/references/output-contract.md
Typical agent-driven usage:
- Provide a local SVG file path or an SVG URL.
- Ask the agent to generate the final deliverables.
- Expect these default outputs:
preview.htmlplayer.js
Typical request examples:
- "Turn this SVG into a hand-drawn preview page."
- "Generate a preview.html and player.js from this SVG."
- "Apply a hand-drawn path and fill animation to this SVG."
A concrete example is included in:
examples/svg-player-demo/975549_OGFB9T1.svgexamples/svg-player-demo/svg-player-demo.html
This example is useful for showing the intended result shape of the skill output and for quick manual inspection in a browser.
By default, the agent should generate:
preview.htmlplayer.js
preview.html is for direct preview.
player.js is for embedding the playback core into another webpage.
<div id="svg-player"></div>
<script src="./player.js"></script>
<script>
createSvgHanddrawPlayer("#svg-player", {
svgMarkup: "<svg>...</svg>",
speed: 1
});
</script>Current player API:
play()pause()seek(ratio)setSpeed(value)destroy()
- This repository keeps the skill independent from any project-specific demo page.
- The intended workflow is agent-driven output generation.
- End users should not need Python, Node, or another local build step unless a self-serve workflow is added later.
- The example demo is included for quick manual inspection, not as a required runtime dependency of the skill.
