demo-preview-environments-manager VS atmos

You can use the setup that is described here: https://github.com/spacelift-io/demo-preview-environments-manager. There's a detailed demo video in the README.

We obviously don't have a project for your exact use case, but we have an open-source example repo that shows a fairly advanced scenario of using the Terraform Spacelift Provider https://github.com/spacelift-io/demo-preview-environments-manager, a simple quickstart of using it https://github.com/spacelift-io/terraform-starter and you can also see the CloudPosse Atmos project, for a very advanced scenario which generates lot's of Stacks based on your component specifications https://github.com/cloudposse/atmos.

Service - AWS Lambda representing our application code, the resulting artifact is a zipfile in S3. Infra - Sets up all the AWS resources the AWS Lambda needs, including the Lambda itself as well as API Gateway resources. Manager - Manages resources representing preview environments. Acts as a repository for all existing preview environments. Setup - Sets up the manager, with all its environment variables and policies. Update File Action - Can be used to create/update/delete files in a different repository. The Service and Infra repository workflows use this to create preview environments in the Manager repository.

And here you can find the repository, which has it all described in its README and is part of the whole setup (the whole demo setup is available publicly on our GitHub): https://github.com/spacelift-io/demo-preview-environments-manager

This is why we created atmos to make it easier to manage large multi-account architectures. As a result, our components are reusable across organizations, regardless of how many accounts and regions they operate, and we minimize the snowflakes. And we avoid code generation, which is hard to thoroughly test in an automated fashion. Without naming names, lots of tools for terraform rely on code generation, but I see it as an anti-pattern that should be avoided.

sorry did not find anything advanced. A better tool to make terraform scaleable is https://atmos.tools

Workspaces are incredibly practical, and we leveraging them at-scale with literally thousands of workspaces using atmos for terraform. There is so much FUD around workspaces that is either ill-informed or based on outdated information. Any company using terraform at scale will rely on tooling and conventions. It's up to that tooling to ensure you are using terraform safely. Atmos is one of those tools. I'm not saying that you have to use workspaces, but just that there's nothing wrong with workspaces themselves.

Check out atmos for a fresh take at managing terraform configurations and terraform workflows. Instead of managing HCL `.tfvar` files manually for configuration, it uses YAML, and supports concepts of imports (via deep merging), remote imports (anything supported by gogetter), mixins, inheritance, multiple-inheritance, vendoring of root modules, workflows, task runners (via custom subcommands), and much more. There's a bit of a learning curve and mind-shift required if coming from a Terragrunt background, but the experience is mindblowing after switching to it. Also, it's not limited to terraform.

We obviously don't have a project for your exact use case, but we have an open-source example repo that shows a fairly advanced scenario of using the Terraform Spacelift Provider https://github.com/spacelift-io/demo-preview-environments-manager, a simple quickstart of using it https://github.com/spacelift-io/terraform-starter and you can also see the CloudPosse Atmos project, for a very advanced scenario which generates lot's of Stacks based on your component specifications https://github.com/cloudposse/atmos.