Are the base image tags current?
They were reviewed on 2026-11-18 and are editable. Check Docker Hub for the newest supported tag of your runtime before you ship, because this page works offline and cannot look tags up.
Small images, non-root, explained.
Pick a stack and get a multi-stage Dockerfile with a pinned base image, a non-root user, a working healthcheck, and a matching .dockerignore, with every line explained. Paste any Dockerfile to lint it with line numbers.
Loading the toolbox.
You do not have to take our word that your data stays here. Developer tools show every request a page makes, so you can watch this page work and confirm that nothing is sent. The check takes under a minute.
F12, or Cmd+Option+I on a Mac, and select the Network tab.Only two things can add a line to that log, and neither carries your data. Submitting the feedback form sends the vote and comment you typed. Typing characters in a script the page has not displayed yet, such as Cyrillic or Greek, can make the browser fetch the matching font file, and that request names only the font.
The production site also sends a Content-Security-Policy header that tells your browser to refuse connections to anywhere except this site and the anonymous analytics endpoint, so a coding mistake could not quietly send your data elsewhere.
It writes a production-shaped Dockerfile for four common stacks, plus the .dockerignore that goes with it and the two commands to build and run the image. Each file follows the same pattern: a build stage that has the compilers and dev dependencies, and a small runtime stage that receives only what the app needs to run.
node user. npm, pnpm, and Yarn are supported, each with its lockfile-respecting install command.CGO_ENABLED=0, copied into a distroless image that has no shell and runs as a non-root user.A single-stage image carries everything that was needed to build the app: compilers, dev dependencies, package caches, and sometimes source files you never meant to ship. All of it adds size, start-up time, and attack surface. A multi-stage build uses one image to build and a second, clean image to run, and copies only the finished output across with COPY --from.
The order of steps matters as much as the stages. Each generated file copies the dependency manifest and lockfile first, installs, and only then copies the rest of the source. Docker reuses a cached layer until one of its inputs changes, so a code change re-runs only the steps after the source copy, and the slow dependency install comes straight from cache.
By default a container process runs as root. If an attacker gets code execution in the app, root inside the container makes every next step easier: writing to the app's own files, installing tools, and exploiting a kernel or runtime bug to escape. Running as an ordinary user removes that head start at no cost.
The generated files use the user each image already provides where there is one (node in the Node images, UID 65532 in distroless :nonroot, UID 101 in nginx-unprivileged) and create a system user with no home directory and no login shell for Python. Application files are copied as root, so the running app can read its code but cannot rewrite it. Root in an earlier build stage is fine and normal; only the final stage matters.
FROM node with no tag means node:latest, which changes whenever a new major version ships, so the same Dockerfile can build a different image next month. Every generated FROM line carries a version tag, and you can edit it. Tags can still be re-pushed with patched contents; for byte-for-byte repeatable builds, add the digest too, as in image:tag@sha256:.... This page works offline and cannot look digests up, so it tells you the command to fetch one instead.
A HEALTHCHECK tells Docker how to ask the app whether it is working. Slim images often have no curl or wget, so a healthcheck that calls curl fails forever and marks a healthy container unhealthy. The generator uses what is really in each image: Node's own http module, Python's urllib, and BusyBox wget on Alpine. Distroless images have no shell at all, so the Go file leaves HEALTHCHECK out and says to use your orchestrator's HTTP probe instead.
Paste any Dockerfile and the linter reports findings with a line and column, grouped as errors, warnings, and notes. It joins continuation lines, skips comments and parser directives, and does not treat heredoc bodies as instructions. The rules include:
:latest base images, and a final stage that runs as root or never sets USER.ENV values or ARG defaults, where docker history shows them to anyone who can pull the image.apt-get install without --no-install-recommends, package lists left in the layer, and apt-get upgrade inside the build.COPY . . before the dependency install, npm install instead of npm ci, and pip without --no-cache-dir.ADD where COPY would do, curl | sh, sudo, cd in RUN, the deprecated MAINTAINER, invalid EXPOSE ports, and unknown instructions.Each rule and its reason are listed in the fold under the linter. The generated files lint clean, which is a useful check after you edit one.
docker build -t app . and then the docker run command shown on the page, and open the mapped port.docker ps once the start period has passed.Running it as a service on a plain Linux host instead? The systemd tool writes the unit file. Keep the .env file you use at run time out of the image with the .dockerignore, and check it with the .env validator. The repository side is covered by .gitignore.
Elsewhere in the toolbox: UUID, .gitignore, CSP, .htaccess, and more on the 2toolbox home page. Part of the 2suite.io privacy-first ecosystem.
They were reviewed on 2026-11-18 and are editable. Check Docker Hub for the newest supported tag of your runtime before you ship, because this page works offline and cannot look tags up.
Distroless images have no shell, curl, or wget, so there is nothing to run a check with. Use your orchestrator probe, such as a Kubernetes liveness probe, instead.
No. Linting runs in this tab, and nothing you paste is sent or stored. You can confirm it in the Network tab.
No. The linter checks common mistakes in the Dockerfile itself. Scan the built image for vulnerable packages and keep the base image updated as well.
npm ci installs exactly what package-lock.json lists and fails if the lockfile is out of date, so every build gets the same dependencies.