Docs

Every post arrives in three forms. Pick the one that matches how much control you want; most sites want the first.

post.html — finished markup

Pre-rendered semantic HTML with syntax highlighting already applied. No renderer to ship, nothing to hydrate, and it works in any framework.

import { RillArticle } from '@rillpage/sdk/article'
import '@rillpage/sdk/styles.css'

export default async function PostPage({ params }) {
  const post = await rill.posts.get((await params).slug)
  if (!post) notFound()

  return (
    <article>
      <h1>{post.title}</h1>
      {post.subtitle && <p>{post.subtitle}</p>}
      <RillArticle post={post} />
    </article>
  )
}

RillArticle renders the cover image, the content, and optional JSON-LD. It deliberately does not render the title or byline — those belong to your page template, where your own typography applies.

Doing it by hand

<div
  className="rill-prose"
  dangerouslySetInnerHTML={{ __html: post.html }}
/>

That is safe here, and it is worth knowing why. post.html is not author-supplied markup: Rill generates it from a document constrained to a fixed schema — headings, lists, tables, code, images, links. No <script> can reach that string, so sanitising it again would be theatre.

post.doc — your own components

The ProseMirror document, if you want a custom code block with a copy button, or images in a lightbox. Walk doc.content and map node types to components. More work; total control.

post.markdown — round-tripping

What the MCP server reads and writes. Useful for exports and migrations; not intended for display.