This repo builds a custom WSL rootfs (tarball) locally or via CI (CodeBuild/GitHub Actions) and uploads it to S3 for distribution.

This custom WSL image comes pre-installed with a comprehensive suite of development tools:

• Node.js: v22 (via nvm)
• Go: v1.23.9
• Java: OpenJDK 17 (or 11 fallback)
• Rust: Installed via rustup
• Python: Python 3 with pip and venv
• C/C++: build-essential, gcc, g++, clang, make

• AWS: AWS CLI v2
• Azure: Azure CLI
• Docker: Docker CLI (docker.io)
• Terraform: HashiCorp Terraform
• Kubernetes:
  • kubectl (Latest stable)
  • helm (Helm 3)
  • kustomize
  • kind (Kubernetes in Docker)
  • minikube
  • k9s (Terminal UI for K8s)
• Security: trivy (Container security scanner)

• Package Managers: npm, yarn, pnpm
• Bundlers/Tools: typescript, vite, webpack, webpack-cli
• IDE: code-server (VS Code in the browser)

• Git: git and gh (GitHub CLI)
• Data Processing: jq (JSON), yq (YAML)
• Modern Unix Tools: fzf (Fuzzy finder), ripgrep (rg), bat (cat clone)
• Document Processing: texlive-full (LaTeX for typesetting documents)
• System: htop, curl, wget, unzip, zip, net-tools, iproute2, procps

On Ubuntu (or WSL host with sudo):

1. sudo apt-get update
2. sudo apt-get install -y debootstrap tar xz-utils sudo
3. sudo bash build.sh
4. Copy custom-wsl.tar.gz to Windows and import:
   • use windows/install-wsl-image.ps1 or: wsl --import MyDevDistro C:\WSL\MyDevDistro C:\path\to\custom-wsl.tar.gz --version 2

Optional: enable systemd by adding [boot]\nsystemd=true to /etc/wsl.conf inside the distro and run wsl --shutdown.

Option 1 — CloudFormation:

1. Push repo to GitHub.
2. Deploy CloudFormation: provide GitHubOwner, GitHubRepo, GitHubBranch, GitHubToken, and optionally ReleaseBucketName.

aws cloudformation deploy --template-file cloudformation/pipeline.yml --stack-name wsl-image-pipeline \
  --parameter-overrides GitHubOwner=your-github-username GitHubRepo=your-repo-name GitHubBranch=main GitHubToken=ghp_YOUR_TOKEN ReleaseBucketName=my-wsl-images

3. Pipeline will produce custom-wsl-<DATE>-<BUILD>.tar.gz and custom-wsl-latest.tar.gz in the Release Bucket (if specified).

Option 2 — GitHub Actions (OIDC):

1. Deploy the OIDC setup stack to create the S3 bucket and IAM role:

   aws cloudformation deploy --template-file cloudformation/github-oidc-setup.yml --stack-name wsl-oidc-setup \
     --capabilities CAPABILITY_NAMED_IAM --parameter-overrides GitHubOrg=your-github-org GitHubRepo=your-repo-name

2. Note the RoleArn and BucketName from the stack outputs. You can retrieve these values using the AWS CLI:

   aws cloudformation describe-stacks --stack-name wsl-oidc-setup --query 'Stacks[0].Outputs'

3. Add the following secrets to your GitHub repository:
   • AWS_ROLE_ARN: The Role ARN from step 2.
   • S3_BUCKET: The Bucket Name from step 2.
   • AWS_REGION: Your AWS region (e.g., us-east-1).
4. Push to main to trigger the build. The workflow uses OIDC for secure, temporary credentials.

Use windows/import-and-setup.ps1 to download & import. You can specify a version (default is "latest").

.\windows\import-and-setup.ps1 -BucketName "my-wsl-images" -Version "latest" -DistroName "MyDevDistro"

• Create a packages.txt file in the root to override the default package list.

• No VPC needed for typical builds.
• Image sizes will be large if you include everything; consider a minimal variant.
• Add IAM least-privilege for production.