Soft question: do you guys prefer using inline or local program execution (or both, of course!) when working with the automation API? I am driving some projects with the automation API and going back and forth on how to structure them, and am trying to stay consistent. Currently each of my subprojects is a local Pulumi program, omitting inline execution entirely. Kind of trivial but would love to hear any thoughts on this (appreciate Pulumi's flexibility in that regard in the first place honestly)

I use inline programs in Go mainly because of the ability to bundle everything as a binary (example: https://github.com/jaxxstorm/connectme) For anything that doesn’t need a bundled binary I use local programs

That's a great point!! Thanks for your response 👍

I guess 1 benefit of not using inline programs is that you can drive that program using either automation API or the CLI - whereas with inline programs you can only use automation API to execute them

Awesome @red-match-15116, thanks for your input. That's what I was thinking too. No joke, almost DM'd you with your deep experience of the automation API haha. Thanks for all the work you put into it/the automation examples/etc.!

Thanks for the kind words @rich-dress-42878!

It is possible to have the best of both worlds with a little bit of refactoring. We don't have an example of this for nodejs, but here is a program in go that can be run both via the CLI and via inline program: https://github.com/pulumi/automation-api-examples/tree/main/go/inline_local_hybrid

Thanks @lemon-agent-27707! I was originally doing that with my current project since I couldn't decide between the two, now switched to pure inline programs. The Go examples have been very helpful and I love how flexible the automation API is/how many ways you can go about it