My Open Knowledge library (static content)

This repository IS the library: an Open Knowledge instance in static content mode reads everything it serves from here. Editing these files and pushing is publishing — no build, no deploy, no database. AGENTS.md documents every format in detail, and doubles as instructions for AI coding assistants (Claude Code, Codex, OpenCode…), so one can assist you in drafting, structuring and publishing your material directly in this repository.

Publish the library in three steps

1. Push this folder to a public GitHub repository (the scaffolder already ran git init and made the first commit):

   git remote add origin git@github.com:<user>/<repo>.git
   git push -u origin main

2. Deploy the Open Knowledge app (once) with the environment variable OK_CONTENT_REPO pointing at this repository's raw URL: https://raw.githubusercontent.com/<user>/<repo>/main

3. Edit, commit, push to publish content from now on. Changes appear within a minute.

Deploy option A — Vercel (no server at all)

The app is stateless in this mode, so a serverless platform works. Click:

https://vercel.com/new/clone?repository-url=https%3A%2F%2Fgithub.com%2Foegea%2Fopen-knowledge&env=OK_CONTENT_REPO&envDescription=Raw%20base%20URL%20of%20your%20content%20repository

…and when asked, set OK_CONTENT_REPO to your raw URL from step 2.

Deploy option B — any machine with Docker

git clone https://github.com/oegea/open-knowledge.git && cd open-knowledge
docker build -t open-knowledge .
docker run -d --name open-knowledge --restart unless-stopped -p 3000:3000 \
  -e OK_CONTENT_REPO=https://raw.githubusercontent.com/<user>/<repo>/main \
  open-knowledge

Deploy option C — container platforms

The same image runs on Railway, Render, Fly.io and friends. No volume is needed — the container is stateless and disposable.

Structure

One file per thing; long-form text lives in Markdown, never inside JSON:

Path	What it is
settings.json	Library name, hero texts, logos, news toggle
courses/index.json	Course directory names, in catalog order
courses/<name>/course.json	One course: metadata, sections, materials, exams
courses/<name>/materials/*.md	One Markdown file per text lesson
categories/<name>.json	Optional card image per category, listed in categories/index.json
news/<name>.json + .md	One news post each, listed in news/index.json
pages/<name>.json + .md	One auxiliary page each, listed in pages/index.json
media/	Images and files, referenced as media/<file>

In this mode there are no accounts: visitors study anonymously and their progress lives in their own browser.