Ran Benita

The change looks good and clean to me, thanks! I think we need to make sure we agree on the semantics we want though, I will post a comment on...

> Scope note: This PR intentionally does not handle nested ExceptionGroups (EGs containing EGs). That can be discussed in a follow-up if desired. Should be fine in a follow up,...

We actually don't expose `Node`, but it's mostly equivalent to `Item | Collector`. However, the reason we don't type `request.node` is that returning a general `Item | Collector` would often...

Sounds good to me. I checked the deploy workflow and I think it should be OK, though we might want to do the github release in two steps - draft...

I checked and `gh release create` uses a draft to upload assets, so it should be good: https://github.com/cli/cli/blob/c0d5f164f20b66b314d71a9ef2a01573e0c156f0/pkg/cmd/release/create/create.go#L508-L510 > I think we should enable this for pytest, or even perhaps...

The main thing I think might break is creating the release and then uploading files to it afterwards. pytest doesn't do it but other projects in pytest-dev might.

Maybe if we announce it first. I'm not sure how to do it, maybe we can enable "Discussions" in the github organization level and announce it there? Or maybe scan...

> FTR I don't think many plugins use GitHub releases -- I have the impression most just publish to PyPI If that's the case it might be OK. --- I...

Thanks for opening the issue. Like @The-Compiler I looked at what are the use cases for `PYTEST_CURRENT_TEST` in the first place. For the "find which test hangs" case, I think...

> I wonder if we should introduce a new "recommended" (but discouraged) way to check whether code is running under testing. Either something like a simple PYTEST=1 env var, or...