Hey everyone 👋 If you’re just getting started with Pulumi and want to convert your existing GCP infrastructure into IaC — we just launched a tool to make that super easy: It’s called StackForge and it does: ✅ Scan any live GCP project ✅ Generate a clean Pulumi

Main.yaml

✅ Output the full Pulumi stack state (

stackName.json

) ✅ No

pulumi import

, no Terraformer, no guesswork 🎥 2-min demo: http://x-itm.com/demo.mp4 🔐 Trial: https://x-itm.com/stackforge-trial/ If you’re starting with an existing GCP environment and want a Pulumi version to edit and deploy — this might save you hours. Happy to answer any beginner questions too 🙏

Output the full Pulumi stack state

Is that by running import internally as part of the tool to build the state file?

Great question Fraser — and no, we don’t run

pulumi import

under the hood. StackForge uses the GCP APIs directly to describe resources and resolve dependencies, then programmatically constructs both the Pulumi YAML (

Main.yaml

) and the associated stack state (

stackName.json

) in Pulumi’s native format. We essentially replicate what you'd get after a successful

pulumi import

, but without the trial/error, manual CLI work, or partial outputs. The result is a clean state file that matches the generated YAML and is deploy-ready — you can open it in the Pulumi Console, run

pulumi preview

, or modify it as needed. Happy to show you an actual example if you want to see how it compares!

I'm surprised building the stack file manually was easier than just running import with an import file. Got to get URN generation and order and all the rest right, but if that worked for y'all then cool.

Totally fair — we were surprised too 😅 The key reason we went this route is that we wanted something anyone could run — even without having Pulumi installed or writing code first. StackForge works from the outside: • No need to scaffold a Pulumi project • No CLI setup, no import loops • No manual ordering or trial/error for dependencies We map resource relationships via GCP APIs, pre-resolve references, and generate fully-linked URNs and YAML in a single pass — including the correct order in the state file. It’s consistent, reproducible, and runs on any GCP project, even remotely or across multiple orgs. We built it specifically so consultants or platform teams could drop into any GCP client environment and get a Pulumi-ready snapshot without touching live infra or running deploys.

That's fair enough, reasonable constraints to have. Come talk to us if you find anything in the CLI/SDKs that would make sense to add to help out with all this.