Applications Overview
An application is the deployable unit in Miabi. It is owned by a workspace and represents a long-running service — a web app, an API, a worker — that Miabi builds, runs, routes, and keeps online for you. You never touch Docker commands; the console handles the full lifecycle.
The three sources
Every application is created from one of three sources:
| Source | What Miabi does | Best for |
|---|---|---|
| Git repository | Clones your repo and builds an image with buildpacks or a detected Dockerfile | Apps you build from source on GitHub, GitLab, or Bitbucket |
| Docker image | Pulls a prebuilt image from a registry | Off-the-shelf images, CI-built artifacts, internal registries |
| Marketplace template | Provisions a curated, versioned template (WordPress, Ghost, n8n, …) | One-click installs of common software |
See Deploy from Git, Deploy from a Docker image, and the Marketplace for each path.
Creating an application
From the workspace dashboard, click New Application, give it a name, and choose a source.
- Name your app — this becomes its identifier within the workspace.
- Pick a source — Git repo, Docker image, or marketplace template.
- Configure the source (repository and branch, image and tag, or template options).
- Set environment variables and resource limits if needed.
- Create — Miabi runs the first build and deploy automatically.
You can create an app with minimal configuration and refine it later. Most settings can be changed and applied on the next deploy.
Application tabs
Once created, each application has a set of tabs in the console:
- Deployments — build and release history, with one-click rollback.
- Environment — environment variables and secrets.
- Domains — connected domains and TLS certificates.
- Logs — live, streaming container logs.
- Settings — source, scaling and resource limits, and lifecycle controls.
What this section covers
The rest of the Applications section walks through each capability: deploying from Git or an image, managing environment variables, releases and rollbacks, scaling, running one-off jobs, grouping apps into stacks, promoting through environments, and watching logs and the timeline.