Skip to content
Back to skills

Ops Docker

ASecurity

Docker and Docker Compose containerization. Trigger when the user wants to dockerize an application or create containers.

  • 5 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added May 28, 2026
ai-agentsbashrailsdockerdebugginggitdatabasesecurity

Security analysis

A100/100

Scanned October 4, 2026

npx -y skills add christopherlouet/claude-base --skill ops-docker --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Ops Docker?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Ops Docker
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/christopherlouet-ops-docker/badge)](https://www.skillsdirectory.com/skills/christopherlouet-ops-docker)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: ops-docker
description: Docker and Docker Compose containerization. Trigger when the user wants to dockerize an application or create containers.
disable-model-invocation: true
---

# Docker Containerization (pointer)

Docker Inc. publishes its own agent skills at [`docker/skills`](https://github.com/docker/skills) (Apache-2.0, pin `v0.3.1`). Four of them cover what this skill used to teach — multi-stage builds, non-root users, `.dockerignore`, `HEALTHCHECK`, BuildKit secrets and cache mounts, Compose healthchecks and `depends_on: service_healthy`:

| Skill | Use it when |
|-------|-------------|
| `docker-project-foundations` | Dockerizing a project, local dev with containers |
| `docker-build-strategies` | Writing, slimming or hardening a Dockerfile |
| `docker-compose-patterns` | Wiring services, adding a database, Compose debugging |
| `docker-destructive-guardrails` | Before any `docker` command that deletes or resets state |

```bash
# pinned release (Docker's README documents the tree/<tag> form)
npx skills add https://github.com/docker/skills/tree/v0.3.1 --skill docker-project-foundations --yes
npx skills add https://github.com/docker/skills/tree/v0.3.1 --skill docker-build-strategies --yes
npx skills add https://github.com/docker/skills/tree/v0.3.1 --skill docker-compose-patterns --yes
npx skills add https://github.com/docker/skills/tree/v0.3.1 --skill docker-destructive-guardrails --yes
```

The repo's other seven skills (Docker Agent, Docker Sandboxes) are product-specific; install them only if you use those products. The three skills that ship a `scripts/verify-*.sh` only build the image and validate the Compose file.

Recipe entry: [`docs/recipes/recommended-vendor-skills.md`](../../../docs/recipes/recommended-vendor-skills.md) §"Docker".

## Foundation discipline the vendor skills leave out

- **Scan the image before it ships**: none of the four mentions vulnerability scanning. Gate the push on `trivy image <image>` (or Docker Scout, Grype) failing on HIGH/CRITICAL — the `ops-ci` skill wires it; `qa-security` owns the policy.
- **Lint the Dockerfile in CI**: `hadolint Dockerfile` as a blocking step, also absent from the vendor skills.
- **Never bake secrets**: no `COPY .env`, no credentials in `ENV` or build args — BuildKit `--mount=type=secret` at build time, orchestrator-injected env at run time (`.claude/rules/security.md`).

## See also

- `/ops:ops-deploy` — deployment checklist consumes the built image
- `/ops:ops-database` — Compose patterns for DB services
- `qa-security` — image scanning gate before push
- `ops-ci` — Hadolint + image scan as CI steps

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…