Visit here
The documentation site for SupaNet - a private AI workspace built on Supabase. Built with Fumadocs (Next.js).
The docs are split for two audiences:
- Using SupaNet (
content/docs/using) - for everyone. What it does and how to get value out of it. No code. - Building on SupaNet (
content/docs/building) - for builders. How it works and how to run and extend it.
npm install
npm run dev # http://localhost:3000
npm run build # production build (typechecks + generates pages)Note: install
node_moduleslocally - they are gitignored and are not checked in. Node 22+ is required.
| Path | What |
|---|---|
content/docs/ |
All the doc pages (.mdx) and section ordering (meta.json) |
app/(home)/page.tsx |
The landing page |
app/docs/ |
The docs layout and dynamic page route |
lib/shared.ts |
App name + GitHub repo config |
lib/layout.shared.tsx |
Shared layout options (nav, GitHub link) |
source.config.ts |
Fumadocs MDX config (frontmatter / meta schemas) |
Add a .mdx file under content/docs/using or content/docs/building, then add
its slug to that folder's meta.json to place it in the nav. Frontmatter needs a
title and description.
This build also emits machine-readable copies of every page:
/<page>- the human page/llms.txtand/llms-full.txt- LLM-friendly summaries of the whole site/llms.mdx/docs/<page>/content.md- the raw markdown of a single page
These make it easy to push the docs into SupaNet's own knowledge base so the in-app assistant can answer from them. See the page Building → Teach SupaNet about itself for the three ways to do it (PDF upload, artifact publish, or MCP push).