Use Docker as a Portable Runtime for One-Shot Commands

Most developers install dozens of CLIs, interpreters, and one-off utilities on their machines over time - jq for JSON, a specific Python version for one script, BusyBox tools to probe a network, a Node binary just to run a build. Each of these adds drift, version conflicts, and "works on my machine" debt. Docker offers a cleaner alternative: treat every container image as a portable, throwaway runtime. Need jq? Run a jq container. Need Python 3.12 for a single script? Run python:3-slim for ten seconds and let it disappear. The host stays minimal; the tooling stays per-task and nothing accumulates.

This mental shift from "containers are how I deploy services" to "containers are how I run commands" is what unlocks Docker's everyday productivity value, especially for engineers who already use docker run daily for servers but never thought to use it for shell utilities. The --rm flag, stdin/stdout piping, and bind mounts (-v $PWD:/work) together make docker run a drop-in replacement for apt install or brew install whenever you need a tool just once. In this lab, you will use four prebuilt images - Alpine, BusyBox, jq, and Python - as one-shot CLIs, pipe data through them, mount host files into them, and confirm that nothing lingers on your machine when each command finishes.

Objectives

Upon completion of this beginner level lab, you will be able to:

• Run a one-shot container with --rm and verify nothing persists in docker ps -a after exit.
• Invoke a network utility from a BusyBox container as if it were a locally installed CLI.
• Bind-mount the host working directory into a container so it can read and write project files.
• Pipe host stdin through a jq container and treat it as a portable JSON command-line processor.
• Run a Python script via python:3-slim without installing Python or any package on the host VM.

Who is this lab for?

This lab is designed for:

• Developers who already know docker pull and docker run and want to use Docker as an everyday command-line utility.
• Platform engineers who want a clean dev box and reach for containers instead of installing tools.
• SREs and ops engineers who want a portable kit of probes (network, JSON, scripting) that runs anywhere Docker runs.