The concept fixture recorded each prompt's last use with the database's wall
clock at seeding, while the page computed "today" from its own clock at
render. A run seeded at 23:59:5x UTC and rendered after midnight read "Used
yesterday" against the concept's "Used today" (PR #284, frontend coverage
shard 2/3).
The concept fixture now takes one instant (optional "now" through the
concept reset hook, the wall clock otherwise) for every last use, mtime and
run time it writes and returns it as manifest.now; seedConceptAt fixes the
browser's Date at that instant. The Library and Images concept specs seed
through it, and the two workspace specs that assert "Used today" and the
"Today" caption fix the page clock at the time they patch or seed. A new
test starts the browser at 23:59:58 UTC, seeds at that instant and passes
midnight before either document opens.
* build(workspace): vendor the approved concept stylesheet with a drift check
* refactor(workspace): adopt the concept stylesheet as the workspace base
* feat(workspace): concept sidebar with real navigation counts
* feat(workspace): concept topbar with Theme and ComfyUI link
* refactor(workspace): one concept heading and footer for every route
* fix(workspace): keep the canvas link on the proxy mount
A resolver failure no longer stops the shell from mounting: the ComfyUI link falls back to the root read from the page's own mount path (never another origin). Tests cover a non-standard /ComfyBackendDirect mount, including that no request escapes it.
* test(workspace): concept fixture and shell comparison
* test(workspace): concept fixture carries the concept inspector facts
* fix(workspace): match the concept brand name and compare shell hover states
* test(workspace): shell comparison mutation check and docs
Add mutate() to the shared concept inventory and a shell mutation check: a
nav-row min-height rule and a removed .breadcrumbs each make the comparison
fail with the changed path. Harden the resolver-failure guard test with an
injection count, make the overlay lint fail when workspace.css is missing,
mark the library.css .row-actions leftovers for PR E, and document the
concept CSS base, the overlay lint, the comparison method and the recorded
shell deviations in TESTING.md.
* fix(workspace): focus the page title on a route change
A sidebar or history navigation now focuses #page-title, as the concept
does on every hashchange (app.js 1168), so a screen reader announces the
page heading instead of the main region. The first render leaves focus
alone and the skip link still focuses main. The heading module's comment
now describes what the shell does.
* fix(workspace): give the Overview nav entry its own home glyph
Overview reused the concept's Activity pulse byte for byte, so once the
Activity row lands two entries would share one icon. Overview now has an
outline home glyph. The shell comparison takes it from the D28 ruling
instead of the concept's activity path, and the icon test checks that
every approved glyph is distinct.
* test(workspace): accept only #249 decision markers and state the lint scope
The overlay lint accepted any D-number, so another issue's ruling such
as "D5 (#245)" could mark a concept override. A marker now has to be a
D-number with no other issue cited; the vocabulary rule that cited #245
reads "#245 decision 5". TESTING.md now states that the lint covers
workspace.css only, lists the unmarked overrides still in images.css,
and says PRs C and D run the lint over each page sheet they keep.
* perf(workspace): re-read only Library and Tags counts after a Library write
A confirmed Library or vocabulary write cannot change the Images count,
but it re-read all three, including the Images output listing. It now
re-reads only the Library and Tags counts. Each count has its own read,
so a write never cancels an Images read that is still in flight.
Vito Sansevero