Docker

Docker Skill

> Skill, Containerization, Docker, Build, Security

Context

Use this skill when writing Dockerfiles, Docker Compose configurations, or optimizing container build pipelines. This covers multi-stage builds, layer caching strategies, .dockerignore, image security scanning, non-root user configuration, and Compose service orchestration for local development. The AI will act as a Docker specialist who understands build performance, image size minimization, and container security fundamentals.

Variables

• {{application_type}}: The application being containerized (e.g., Node.js API, Python FastAPI, Go binary, Java Spring Boot).
• {{base_image_preference}}: Preferred base image family (e.g., alpine, debian-slim, distroless, ubi-minimal).
• {{build_goal}}: What the Dockerfile/Compose needs to accomplish (e.g., production image, local dev environment with hot-reload, multi-arch build).
• {{registry}}: Target container registry (e.g., Docker Hub, AWS ECR, GitHub Container Registry, GitLab Registry).

Prompt

Adopt the persona of a Senior Docker / Containerization Engineer. I need to containerize the following application:

Application: {{application_type}}
Base Image Preference: {{base_image_preference}}
Goal: {{build_goal}}
Target Registry: {{registry}}

Design the Docker configuration adhering to these standards:

1. **Multi-Stage Build:** Use a multi-stage `Dockerfile` with a named `builder` stage and a lean final `runtime` stage. The final image must contain only what is needed to run the application — no build tools, package managers, or source code.
2. **Layer Caching:** Order instructions from least to most frequently changing. Copy dependency manifests (`package.json`, `requirements.txt`, `go.mod`) and install dependencies *before* copying application source code to maximize cache reuse.
3. **Image Size:** Choose the smallest appropriate base image. Combine `RUN` commands with `&&` and clean up caches in the same layer (e.g., `rm -rf /var/cache/apk/*`). Provide a `docker images` size comparison between a naive and optimized approach.
4. **Security:**
   - Run as a non-root user (`USER appuser`). Create the user explicitly with a fixed UID.
   - Use `--no-cache` for package managers in CI.
   - Set `COPY --chown=appuser:appuser` for file ownership.
   - Recommend a scanning step (e.g., Trivy, Docker Scout) for the CI pipeline.
5. **.dockerignore:** Provide a comprehensive `.dockerignore` that excludes `.git`, `node_modules`, test files, `.env` files, and build artifacts.
6. **Docker Compose (if applicable):** For local development, provide a `docker-compose.yml` with named volumes for persistence, `healthcheck:` definitions, and a `.env.example` pattern for environment variables.
7. **Build Arguments & Labels:** Use `ARG` for build-time variables (e.g., `APP_VERSION`). Add `LABEL` metadata (version, maintainer, build date) using OCI standard labels.

Provide the complete `Dockerfile`, `.dockerignore`, and `docker-compose.yml` (if needed) with inline comments.

Example Usage

Input:

Adopt the persona of a Senior Docker / Containerization Engineer. I need to containerize the following application:

Application: Node.js 20 Express API with TypeScript (compiled to dist/)
Base Image Preference: alpine for production, node:20 for dev
Goal: A production-optimized multi-stage build, plus a docker-compose.yml for local development with hot-reload via nodemon
Target Registry: GitHub Container Registry (ghcr.io)

Design the Docker configuration adhering to these standards:
[...rest of prompt...]

Expected Output:

• Multi-stage Dockerfile: builder stage (node:20-alpine, installs all deps, compiles TS), runtime stage (node:20-alpine, copies only dist/ and node_modules/ production deps)
• Final image runs as node user (UID 1000), not root
• .dockerignore excluding src/, *.ts, .git, node_modules, coverage/
• docker-compose.yml with app service mounting src/ as a volume for hot-reload, a postgres service with a named volume, and healthcheck on both services
• OCI labels for versioning

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: DongDuong2001
• Source: DongDuong2001/pudo-code-system
• License: MIT
• Homepage: https://dongduong2001.github.io/pudo-code-system/

Install and usage instructions live in the source repository linked above.