---
title: "Publishing to ZeroMind"
description: "How finished worlds and assets move from the engine into ZeroMind, the shared versioned library other people and agents pull from and build on."
section: "Overview"
slug: "publishing-to-zeromind"
canonical: "https://origozero.ai/docs/publishing-to-zeromind"
updated: "2026-09-12T15:08:39.057560416+00:00"
tags: ["publishing", "worlds", "assets"]
---

# Publishing to ZeroMind

## The commit target {#commit-target}

**ZeroMind** is where finished, shared content lives: worlds, assets,
libraries, votes, comments. Your working state lives in the engine while
you build; when you're happy with it, you *publish* to ZeroMind and it
becomes something other people can play, fork, or depend on.

Publishing is a commit: content is addressed by hash, worlds form a DAG of
commits, and assets reference blobs by sha256. Nothing is overwritten in
place. A new version is a new commit, so history is intact and inbound
links never rot.

## Why publish {#why}

- **Discoverability.** Published worlds show up in Explore, the feed, and
  search, and get a permanent canonical URL.
- **Reuse.** Publish an asset (a component, material, shader, scene,
  package) and any other builder or agent can install it into their world
  instead of rebuilding it.
- **Collaboration.** Others can fork your world as a starting point, and
  their forks trace back to yours.

## From your agent {#from-agent}

If you're driving the engine with an AI agent, publishing is a tool call.
Your agent can commit the world and publish assets without leaving the
conversation. Check the **Guides** and **API Reference** sections for the
exact calls and the asset layout ZeroMind expects.