magic_starter 0.0.1-alpha.20

[anilcancakir]

Version bump and changelog date. Two files, matching #90's shape.

What ships

Exactly two commits since 0.0.1-alpha.19, both about one defect:

• Let the theme turn a page header inline, so its two halves cannot drift #92 — MSPageHeader's inline mode is settable from the theme, so its two halves cannot drift apart.
• Document the page header's inline mode, which the last change skipped #93 — the docs and CHANGELOG that Let the theme turn a page header inline, so its two halves cannot drift #92 skipped.

Nothing else is in [Unreleased], so this release is the header fix and its paper, and nothing rides along.

Why it is worth a release rather than waiting

inlineActions does two things at once: it selects containerInlineClassName, and it gives the title row flex-1 min-w-0 instead of sm:flex-1. The container class is a theme string and the flex behaviour was a widget argument, so a consumer could set the first and not the second, and that combination silently overflows. MSPageScaffold does not forward the argument at all, so a scaffold consumer could not reach the second half even knowing it existed.

The consumer that hit it is depools, whose product screen reports A RenderFlex overflowed by 40 pixels on the right at 390 today. It cannot take the fix until this is on pub.dev: it pins ^0.0.1-alpha.17, and the field does not exist in alpha.19.

Not breaking

MSPageHeader.inlineActions became bool? falling back to a theme field that defaults to false, which is the previous behaviour, and an explicit argument still wins. Every existing caller is unchanged.

Verification

dart pub publish --dry-run clean apart from the two files this PR is about (they read as modified until it is committed) and the standing pubspec_overrides.yaml hints, which are the workspace's local sibling paths and are gitignored, so CI resolves the hosted versions instead.

dart format --set-exit-if-changed . clean, dart analyze clean, 1253 tests pass. The overflow is one of them: reverting the title-row half reports A RenderFlex overflowed by 79 pixels on the right.

After merge

git tag 0.0.1-alpha.20 && git push --tags fires publish.yml, which is the trigger this repo uses.

One thing noticed and left alone

CLAUDE.md's banner still reads **Version:** 0.0.1-alpha.14, six releases behind. Previous release PRs did not touch it either, so it is clearly not part of this ritual, and correcting it inside a release PR would add review surface with no release value. Worth its own one-line PR.

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!