Add reusable workflow for checking build output

[snoack]

This extracts the shared build action used by work-diary and update-changelog-action into a reusable workflow.

---

This change is 

[pkaminski]

@pkaminski reviewed 1 file and all commit messages, and made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on snoack).

---

.github/workflows/check-build-output.yaml line 18 at r1 (raw file):

        with:
          node-version: "24"
          cache: yarn

At least Codex was convinced that actions/setup-node was incapable of caching Yarn 4 correctly, and explicit use of actions/cache was required. Are you sure this works? If so, should backport to a bunch of other workflows.

[snoack]

@snoack made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on pkaminski).

---

.github/workflows/check-build-output.yaml line 18 at r1 (raw file):

Previously, pkaminski (Piotr Kaminski) wrote…

At least Codex was convinced that actions/setup-node was incapable of caching Yarn 4 correctly, and explicit use of actions/cache was required. Are you sure this works? If so, should backport to a bunch of other workflows.

Well, Codex came up with this simplification, and asking it about it again now, it insists that this works with Yarn 4 since actions/setup-node internally uses actions/cache. I did no test it. I suppose once this has been merged we can rerun the CI for the integration changes in work-diary and update-changelog-action and see if it works.

[pkaminski]

@pkaminski made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on snoack).

---

.github/workflows/check-build-output.yaml line 18 at r1 (raw file):

Previously, snoack (Sebastian Noack) wrote…

Well, Codex came up with this simplification, and asking it about it again now, it insists that this works with Yarn 4 since actions/setup-node internally uses actions/cache. I did no test it. I suppose once this has been merged we can rerun the CI for the integration changes in work-diary and update-changelog-action and see if it works.

How would we know it's working, though?

The Yarn maintainers seem to think (or did, 5 years ago) that caching node_modules and .yarn/install-state.gz would often be beneficial for our configuration.