DevOps Engineer Skill for Claude Code: CI/CD, Docker, Kubernetes, Terraform
A skill that makes your agent work like a senior DevOps engineer, and ask for your approval before anything goes to production.
- Installs
- 11.5K
- on skills.sh
- Stars
- 12.2K
- on GitHub
- Files
- 9
- besides SKILL.md
- Updated
- Oct 2026
- last repo commit
The devops-engineer skill gives your agent a fixed workflow for infrastructure work: assess, design, implement, validate, plan the rollout, approve and deploy, then monitor. It covers GitHub Actions, GitLab CI, Docker, Kubernetes, Terraform, deployment strategies, platform engineering, release automation and incident response, each in its own reference guide that loads when relevant.
The rules are strict about production. SKILL.md says to run terraform plan and confirm no destructive changes first. For production or customer-facing targets the agent must show a deployment summary and rollback plan, ask for explicit approval, and stop if you do not give it. The skill bundles no scripts, but the agent runs commands in your shell.
Good for
- Developers who want a working CI/CD pipeline or Dockerfile for their app
- Teams writing Kubernetes manifests or Terraform and wanting a validation and rollback habit
- On-call engineers who want incident runbooks and postmortem templates
Install devops-engineer
Claude Code
- 1
Copy this prompt.
Install the devops-engineer skill in Claude Code. Run this command: /plugin marketplace add jeffallan/claude-skills /plugin install fullstack-dev-skills@fullstack-dev-skills Then check that it loaded and tell me how to use it.
- 2
Paste it into Claude Code. It shows the change and asks before running it.
The plugin installs all 67 skills and also its slash commands, which the README says need an Atlassian MCP server. The skills CLI route installs only devops-engineer. Pick one route, not both.
From the project README. Source.
What's in the skill
When it loads
Creates Dockerfiles, configures CI/CD pipelines, writes Kubernetes manifests, and generates Terraform/Pulumi infrastructure templates. Handles deployment automation, GitOps configuration, incident response runbooks, and internal developer platform tooling. Use when setting up CI/CD pipelines, containerizing applications, managing infrastructure as code, deploying to Kubernetes clusters, configuring cloud platforms, automating releases, or responding to production incidents. Invoke for pipelines, Docker, Kubernetes, GitOps, Terraform, GitHub Actions, on-call, or platform engineering.
Sections in SKILL.md
- Role Definition
- When to Use This Skill
- Core Workflow
- Reference Guide
- Constraints
- MUST DO
- MUST NOT DO
- Output Templates
- Minimal GitHub Actions Example
- Minimal Dockerfile Example
- Rollback Procedure Example
- Knowledge Reference
Key facts
- Installs on skills.sh
- 11,520 as of 2026-11-02
- License
- MIT
- Last repo commit
- 2026-10-30
What it includes
Assess, design, implement, validate, plan rollout, approve and deploy, monitor.
Validation runs terraform plan, lints configs and runs tests. The deploy step needs your explicit approval for production.
Runs code ·
Seven-step core workflowShows a deployment summary and rollback plan, then waits for you.
If you withhold approval, the skill says to stop with a blocked verdict.
Read ·
Approval gate for productionReference for CI/CD workflows.
Loaded when setting up pipelines or GitHub workflows.
Write ·
GitHub Actions guideReference for .gitlab-ci.yml, DAG needs, environments and runners.
Loaded for GitLab pipelines.
Write ·
GitLab CI/CD guideReference for Dockerfiles and Compose.
Loaded when containerizing an app.
Write ·
Docker guideReference for deployments, services, ingress and pods.
Its examples include kubectl apply.
Runs code ·
Kubernetes guideReference for infrastructure as code on AWS and GCP.
Its examples include terraform apply and terraform destroy.
Runs code ·
Terraform guideBlue-green, canary, rolling updates and rollback.
Includes a rollback procedure using kubectl rollout undo.
Runs code ·
Deployment strategies guideSelf-service infrastructure, developer portals, golden paths and Backstage.
Loaded for internal platform work.
Write ·
Platform engineering guideArtifact management, feature flags and multi-platform CI/CD.
Examples include docker push and kubectl apply.
Runs code ·
Release automation guideOutages, on-call, MTTR, postmortems and runbooks.
Examples include kubectl delete pod in production.
Runs code ·
Incident response guide
From the official docs
Permissions
Medium risk
The agent runs infrastructure commands in your shell, and the guides include ones that apply, destroy or push things. Try it on a test account or cluster before real environments.
What it can do
- Read your data
- Create and change things
- Delete things
- Read-only mode
Watch out for
Examples can destroy infrastructure
terraform-iac.md shows terraform apply and terraform destroy, and incident-response.md shows kubectl delete pod in production. Run these only against environments you can rebuild. github.com
Pushes images and applies manifests
release-automation.md has docker push and kubectl apply steps. Check the registry and cluster before approving. github.com
The plugin route adds 67 skills and slash commands
The README says the workflow commands need an Atlassian MCP server. The skills CLI route installs just this skill. github.com
Technical details
| Turning on read-only | No switch. SKILL.md requires explicit approval before production deployment commands. |
|---|---|
| Sign-in and keys | SKILL.md names no key. Commands use the cloud, cluster and registry sign-ins already on your machine. |
| Scopes |
|
Questions
What does the devops-engineer skill do?
It makes your agent act as a senior DevOps engineer. It follows a seven-step workflow for pipelines, containers, Kubernetes, Terraform and incident response, and asks for your approval before production deploys.
How do I install it?
Use the install box at the top of this page. The plugin route installs all 67 skills in the repo (counted from the plugin branch tree) plus workflow slash commands that the README says need an Atlassian MCP server. The skills CLI route installs only this skill. There is no Claude app upload route here because its guides run terraform, kubectl and docker on your own machine.
Will it deploy to production on its own?
No. SKILL.md says to present a deployment summary and rollback plan, ask for explicit approval, and run deployment commands only after confirmation.
What does it refuse to do?
Its MUST NOT list: deploy to production without approval, store secrets in code or CI/CD variables, skip staging tests, ignore container resource limits, use the latest tag in production, and deploy on Fridays without monitoring.
Does it need an API key?
SKILL.md names none. Anything it runs uses the cloud, cluster and registry sign-ins already on your machine.
Is it safe to use?
It bundles no scripts, but its guides include commands that apply, scale back or destroy real infrastructure. Use a test account or cluster first, and read each command before you approve it.
What other skills does it list?
The frontmatter lists related skills: terraform-engineer, kubernetes-specialist, sre-engineer, monitoring-expert and security-reviewer. They are optional hand-offs, not requirements.
Information
updated Nov 2, 2026
- Repository
- jeffallan/claude-skills
- Folder
skills/devops-engineer
Sources
Docs, issue trackers and discussions this page is built from.
Show all 6Hide
- https://github.com/Jeffallan/claude-skills/blob/HEAD/README.md · github.com
- https://github.com/Jeffallan/claude-skills/blob/HEAD/QUICKSTART.md · github.com
- https://github.com/vercel-labs/skills/blob/main/docs/commands.md · github.com
- https://github.com/jeffallan/claude-skills/blob/HEAD/skills/devops-engineer/SKILL.md · github.com
- https://github.com/Jeffallan/claude-skills/blob/HEAD/skills/devops-engineer/references/terraform-iac.md · github.com
- https://github.com/Jeffallan/claude-skills/blob/HEAD/skills/devops-engineer/references/incident-response.md · github.com