Standalone by default
Clone the repo, edit `story.md`, add a CSV in `public/data`, and render a working scrollytelling page.
A standalone starter for turning story specs and CSV files into scroll-driven webpages.
scrollytale-starter is a standalone project for building scroll-driven data stories with:
story.mdpublic/data/The core idea is simple:
story.mdThis repo is designed for both humans and AI agents.
story.mdnpm install && npm run devUse this repo as the runtime and template layer.
Read in this order:
story.mdDefault behavior:
story.mdpublic/datasrc/styles/custom.css only when neededAvoid changing runtime code unless the user explicitly asks for a new feature.
This repo is:
This repo is not:
structure: linearlayout: chapter, hero, scrolly-left, scrolly-overlay, scrolly-right, full-widthvis.type: html, bar, line, unitThe broader academic design space is documented in design-space.md. The starter already accepts some of that metadata in story.md, even when the runtime support is still partial.
Treat that page as advanced/internal guidance. New stories should usually start from the simpler examples in story-format.md.
This starter is meant to work well with a separate AI authoring skill named scrollytale.
That skill helps an AI:
story.mdBut the starter must remain usable on its own. Someone should be able to clone this repo and manually edit story.md without using any separate skill tooling.
This documentation site is powered by VitePress and intended for:
To publish it, enable GitHub Pages in the repository and choose:
GitHub Actions