Quick Start Guide¶

Get up and running with DevOps Images in five minutes. All you need is Docker (or any OCI runtime, such as Podman).

flowchart LR
  A["1. Choose<br/>an image"] --> B["2. Pull it"] --> C["3. Run it"] --> D["4. Try a tool"] --> E["5. Mount your<br/>project and credentials"]
  classDef s1 fill:#0369a1,stroke:#075985,color:#fff
  classDef s2 fill:#059669,stroke:#047857,color:#fff
  classDef s3 fill:#0d9488,stroke:#0f766e,color:#fff
  classDef s4 fill:#d97706,stroke:#b45309,color:#fff
  classDef s5 fill:#ea7a0c,stroke:#c2410c,color:#fff
  class A s1
  class B s2
  class C s3
  class D s4
  class E s5

Step 1: Choose your image¶

all-devops

AWS and Google Cloud tooling. A good default if you're unsure or work across clouds.
aws-devops

AWS CLI v2, Session Manager, boto3 and cfn-lint. No Google Cloud SDK.
gcp-devops

The Google Cloud CLI, GKE auth plugin and docker-credential-gcr. No AWS tooling.

Still deciding? The interactive image picker asks three questions and gives you the command.

Step 2: Pull the image¶

all-devopsaws-devopsgcp-devops

docker pull ghcr.io/jinalshah/devops/images/all-devops:latest

docker pull ghcr.io/jinalshah/devops/images/aws-devops:latest

docker pull ghcr.io/jinalshah/devops/images/gcp-devops:latest

The first pull takes a minute

Each image is about 1.5–1.6 GB to download and roughly 4.6–5 GB on disk once unpacked. After that, starting a container is instant. Grab a coffee!

Step 3: Run it interactively¶

docker run -it --rm ghcr.io/jinalshah/devops/images/all-devops:latest

You land in Zsh with Oh My Zsh's candy theme. The prompt looks like this:

root@3f2a1b9c8d7e [10:42:07] [/]
-> #

Type exit (or press Ctrl+D) to leave. --rm deletes the container afterwards.

Step 4: Try a tool¶

terraform version
kubectl version --client
aws --version         # all-devops and aws-devops
gcloud --version      # all-devops and gcp-devops
claude --version      # plus codex, copilot and agy

You're in

Every tool prints its version, so everything is installed and on your PATH.

Step 5: Set up for real work¶

Mount your project and the credentials you already have on your machine:

docker run -it --rm \
  -v "$PWD":/srv -w /srv \
  -v ~/.ssh:/root/.ssh:ro \
  -v ~/.aws:/root/.aws \
  -v ~/.config/gcloud:/root/.config/gcloud \
  ghcr.io/jinalshah/devops/images/all-devops:latest

Your project is at /srv, and the cloud CLIs use your existing logins.

Build your command¶

Pick an image and toggle what you want mounted. The command updates as you click, ready to copy.

Mount	What it gives you
`-v "$PWD":/srv -w /srv`	Your project files, as the working directory
`-v ~/.ssh:/root/.ssh:ro`	SSH keys for Git and servers (read-only)
`-v ~/.aws:/root/.aws`	AWS CLI profiles and SSO cache
`-v ~/.config/gcloud:/root/.config/gcloud`	gcloud logins and configurations
`-v ~/.kube:/root/.kube`	kubeconfig for your clusters
`-v ~/.claude:/root/.claude`	Claude Code login and settings
`-v ~/.codex:/root/.codex`	Codex CLI login and `config.toml`
`-v ~/.copilot:/root/.copilot`	Copilot CLI login and settings
`-v ~/.gemini:/root/.gemini`	Antigravity CLI (`agy`) login and settings

What you get¶

Infrastructure as code

Terraform (via tfswitch), Terragrunt, TFLint, Packer
Kubernetes

kubectl, Helm 3, k9s
Automation and security

Ansible, ansible-lint, pre-commit, Task, Trivy
AI coding agents

Claude Code, Codex CLI, Copilot CLI, Antigravity CLI (agy)
Development

Python 3.14, Node.js LTS, Git, GitHub CLI, ghorg, Zensical
Databases and network

mongosh, psql, mysql · dig, nmap, ncat, curl, jq

Cloud tooling per image¶

Tool	aws-devops	gcp-devops
AWS CLI v2 + Session Manager plugin		—
boto3, cfn-lint, s3cmd		—
gcloud, gsutil, bq	—
GKE auth plugin, docker-credential-gcr	—

Search the whole list in the tool explorer.

Common first tasks¶

Terraform Security scan Kubernetes AI review

cat > main.tf <<'EOF'
output "hello" {
  value = "Hello from DevOps Images!"
}
EOF

docker run --rm -v "$PWD":/srv -w /srv \
  ghcr.io/jinalshah/devops/images/all-devops:latest \
  sh -c "terraform init && terraform apply -auto-approve"

# Scan dependencies, secrets and IaC misconfigurations in the current folder
docker run --rm -v "$PWD":/srv -w /srv \
  ghcr.io/jinalshah/devops/images/all-devops:latest \
  trivy fs --scanners vuln,secret,misconfig .

Trivy downloads its vulnerability database on first run.

# Validate manifests against your cluster's API without changing anything
docker run --rm -v "$PWD":/srv -w /srv \
  -v ~/.kube:/root/.kube \
  ghcr.io/jinalshah/devops/images/all-devops:latest \
  kubectl apply --dry-run=server -f k8s/

# One-time: sign in (opens a URL to visit), then /exit
docker run -it --rm -v ~/.claude:/root/.claude \
  ghcr.io/jinalshah/devops/images/all-devops:latest claude

# Review a file non-interactively with -p (print mode)
docker run --rm -v "$PWD":/srv -w /srv \
  -v ~/.claude:/root/.claude \
  ghcr.io/jinalshah/devops/images/all-devops:latest \
  sh -c 'cat main.tf | claude -p "Review this Terraform for security issues"'

See AI CLI setup for Codex, Copilot and Antigravity.

Next steps by use case¶

Local development CI/CD Security-focused

Keep a named container around between sessions:

docker run -it --name devops-work \
  -v "$PWD":/srv -w /srv \
  -v ~/.ssh:/root/.ssh:ro \
  -v ~/.aws:/root/.aws \
  -v ~/.config/gcloud:/root/.config/gcloud \
  -v ~/.claude:/root/.claude \
  ghcr.io/jinalshah/devops/images/all-devops:latest

# Later
docker start -ai devops-work

Then set up cloud credentials and AI CLIs.

container:
  image: ghcr.io/jinalshah/devops/images/all-devops:1.0.abc1234

Pin a tag: 1.0.<short-sha> tags are per commit. Scheduled rebuilds refresh them with newer tools, so pin a @sha256: digest if you need exact bits.
Use CI secrets for credentials, never baked-in files.
Read more in CI/CD integration.

Scan code and IaC with trivy fs and trivy config
Lint with TFLint and ansible-lint
Run pre-commit hooks in CI
Pin image digests
Scan the image itself: trivy image ghcr.io/jinalshah/devops/images/all-devops:latest

See the security-first workflow.

Handy shell aliases¶

Add these to your ~/.bashrc or ~/.zshrc:

# Interactive shell in the current project
alias devops='docker run -it --rm -v "$PWD":/srv -w /srv -v ~/.ssh:/root/.ssh:ro -v ~/.aws:/root/.aws -v ~/.config/gcloud:/root/.config/gcloud ghcr.io/jinalshah/devops/images/all-devops:latest'

# One-off commands: devops-run terraform plan
alias devops-run='docker run --rm -v "$PWD":/srv -w /srv -v ~/.aws:/root/.aws -v ~/.config/gcloud:/root/.config/gcloud ghcr.io/jinalshah/devops/images/all-devops:latest'

Troubleshooting your first run¶

docker: command not found

Install Docker Desktop (macOS, Windows) or Docker Engine on Linux.

Cannot connect to the Docker daemon

Start Docker Desktop, or on Linux run sudo systemctl start docker.

Permission denied while trying to connect (Linux)

sudo usermod -aG docker "$USER"
# Log out and back in for this to take effect

The pull is very slow

It's a 1.5 GB+ download, so the first pull takes a while. Use GHCR rather than Docker Hub, which rate-limits anonymous pulls, pick the single-cloud image if you only need one cloud, and it only happens once per update.

A tool says command not found

Check you're using the right variant: aws isn't in gcp-devops and gcloud isn't in aws-devops. The tool explorer shows which image has what.

More fixes are in Troubleshooting.

Where next?¶

Tool basics

A cheat sheet for every tool.
Authentication

AWS, GCP, SSH, Git and AI CLIs.
AI CLI setup

Claude, Codex, Copilot and Antigravity.
Workflows

CI/CD recipes and multi-tool patterns.