Skip to content

Documentation ​

Shipwick runs Docker applications on a single Linux server of your own: rolling deployments with health checks and rollback, HTTPS, jobs, backups, secrets and a dashboard, from one small deploy.yaml. This page says what it is, what it is not, and where to find things.

I just want to deploy

  1. Install the CLI on your laptop, one command
  2. Install on a server, one command from your laptop
  3. Your first deployment, from init to a URL

You need a Linux server you can reach over SSH, a domain, and your project. shipwick init writes the Dockerfile for a Node, Next.js, Nuxt, SvelteKit, Remix, Astro, .NET, Go or Python project, and a folder of static files needs no container at all. How it fits together is the picture in one page.

I run servers for a living

  1. Architecture: what runs where, what is stored, how the parts talk
  2. Security: the trust model, tokens and sign-in, and how to expose the API
  3. Reference: deploy.yaml, CLI, API, Agent configuration

Every field, command, variable and endpoint is documented; every number in the reference was measured.

What Shipwick is ​

Shipwick is a single agent that runs on your server and turns a deploy.yaml into running, supervised containers.

yaml
# deploy.yaml
name: my-api
build: .
port: 8080
domain: api.example.com
replicas: 2
bash
shipwick deploy

It is for developers and small teams running 1–20 applications on a VPS — Hetzner, DigitalOcean, OVH, EC2 or similar — who want Docker in production without writing their own deploy scripts, restart logic, health checks, rollbacks and reverse-proxy configuration.

  • One binary, one SQLite file. No cluster, no control plane, no external database.
  • Docker is the runtime. Anything that runs with docker run runs on Shipwick. The image is pulled from a registry, or built on your machine by shipwick deploy and sent to the server; no registry is needed for that.
  • A failed deployment never takes down the version that works.

The parts:

PartRole
AgentRuns on the server. Owns the deployment lifecycle, supervises containers, configures Caddy. Keeps its state in SQLite.
shipwickThe command-line client, for your laptop and for CI. Builds images with build: ., uploads static folders, and installs the server over SSH.
DashboardThe same information and everyday actions in a browser.
CaddyServes application domains over HTTPS, and static sites straight from disk. The agent tells it what to route where.

What Shipwick is not ​

Shipwick is for one server, by design. It schedules nothing across machines, runs no service mesh and has no extension model. The agent builds no images: with build: . the build runs on your machine, and the server only loads the result. It needs no external database or queue.

When one server is no longer enough, you have outgrown Shipwick. Until then, it is the whole platform.

Status: 1.0

The current version is 1.1.0. From 1.0 on, deploy.yaml, the API and the files on the server change only with a migration and a documented upgrade, a minor release removes and renames nothing, and a 1.x CLI works against any 1.x agent: Versions and compatibility says what that means in full, and the changelog names every change.

How the documentation is organized ​

Getting started ​

Install the CLI and the server, then deploy an application.

Concepts ​

How Shipwick works and why it behaves the way it does.

Tasks ​

How to do one specific thing.

Deploy and run

Domains and HTTPS

Data and backups

Access

Watch

The server

Reference ​

Every field, command, variable and endpoint.

Security ​

An admin token is equivalent to root SSH access to the server. Read Security before you put Shipwick on a machine that matters. Where your data is collects what a server stores, encrypts, logs and sends, for whoever has to answer a questionnaire about it.

Source and license ​

Shipwick is open source under the Apache License 2.0. The source, the issue tracker and the releases are at github.com/shipwick/shipwick.