sync(upstream): merge upstream/main 5661c6116 into fork/dev

[omegent-app]

Adopts upstream through 5661c6116 — the three commits after #378, as two merge commits
(#5372 landed first; #5581 arrived while this PR was in review and is folded in here).

Important

Merge with a merge commit, not squash. Squashing collapses the second parent and discards
upstream's commits as ancestors — the thing that keeps "commits behind upstream" honest.

Commit
5661c6116	feat(web): drag pinned threads into your own order (pingdotgg#5581)
6d70e6d77	fix(mobile): reconnects no longer shift the thread list (pingdotgg#5372)
be1a83674	chore(release): prepare v0.0.32

What pingdotgg#5372 does, and why it collided

Upstream moves mobile connection status out of the lists into the header title slot (new
WorkspaceConnectionTitle / getConnectionAwareBrandHeaderOptions) and deletes
WorkspaceConnectionStatus.tsx, so a reconnect no longer shifts rows. The fork rendered that status
in four places across HomeScreen and ThreadNavigationSidebar — exactly the row-shifting being
fixed — so those in-list surfaces are removed and the header slots adopted.

The auto-merge left a dangling {connectionStatus} in HomeScreen whose definition upstream had
deleted; typecheck caught it.

Fork surfaces kept around the new title:

• multi-environment selection — HomeScreen keeps its selectedEnvironmentIds label derivation
  and environmentLabelById (BoardScreen consumes it) instead of upstream's single
  selectedEnvironmentId;
• board mode — the sidebar keeps its board branch, board-specific solid header and board-gated
  search bar, with upstream's connection-aware options merged into the same header object (spread
  first, so board's transparency overrides still win);
• HomeRouteScreen keeps the fork's in-flow HomeHeader as title owner rather than also
  restoring upstream's native-stack title.

Adversarial review — two rounds, and it earned its keep

Round 1 (grok-4.5 + gpt-5.6-sol, independently) found two real defects in my first pass:

Finding	Disposition
HIGH (both) — iOS Home has its own IosHomeHeader owning the native title, so dropping upstream's route-level options left no connection-status surface at all on that path	Fixed. IosHomeHeader now spreads the connection-aware options
MED (grok) — getConnectionAwareBrandHeaderOptions hardcodes the brand lockup and the literal "Threads", so spreading it over the fork's list-mode titles silently renamed Board and Projects	Fixed. The helper takes an optional title/brand; every adopting surface passes its mode title
LOW (both) — dead getConnectionAwareBrandHeaderOptions import left in HomeRouteScreen	Fixed. Removed

Round 2 re-verified the fixes and caught a surface I had missed:

Finding	Disposition
MED (both) — the custom (Android split) sidebar large title still hardcoded "Threads"	Fixed. Now mode-aware
LOW (codex) — brandTitleOffset was applied to caller-supplied titles that lack the brand lockup's internal inset, sliding the status sideways during the swap	Fixed. The offset now applies only to the default brand
LOW (grok) — 17px vs 18px title metrics on the new iOS brand text	Fixed. Aligned to 18
(codex) — no test covered mode-title preservation, "which allowed the custom-sidebar defect through"	Fixed. mobileSurfaceExistence now asserts every adopting surface passes its own list-mode title and that no surface re-renders the deleted in-list pill. Verified to fail on the pre-fix tree

Final coverage, confirmed by both reviewers: status is reachable on Android in-flow home, iOS home,
sidebar native, sidebar custom, board mode and split view — exactly once on each, with mode
titles surviving everywhere.

pingdotgg#5581 — drag pinned threads (second merge)

Pinned rows now sort by the user's arranged pinOrderKey on every platform, deliberately
independent of the fork's thread-grouping preference: that preference still orders the active
rows, but a pin is an explicit arrangement, and upstream documents the sort as identical everywhere
so mixed-version fleets cannot disagree. The threadPinReorder capability gates dragging only,
never the sort. SidebarV2 keeps the fork's environment/ownership filters around the new pinned
block and gains reorderablePinnedKeys.

Welds the auto-merge produced, every one caught by typecheck or tests:

Where	Weld
useThreadActions.ts	upstream's ThreadPinReorderUnsupportedError landed inside the fork's clearPerThreadClientState, leaving it unclosed
threadListV2.ts	two for (const thread of …) headers stacked on one loop body
useThreadListActions.ts	duplicate threadEnvironment import plus local copies of helpers this fork imports from threadActionMessages
HomeRouteScreen	upstream's single-environment onEnvironmentChange handler, which this multi-select screen has no prop for
projector.pinned.test.ts	indexed state.threads[0]; this fork's read model keys threads by id in an Effect HashMap
six migration fixtures	pinned the ledger tail at 37; upstream's 038 ProjectionThreadsPinOrderKey extends it

Review of the second merge: both reviewers cleared pin-sort/capability separation, fork
behaviour, welds, ledger windows and test integrity. Two follow-ups:

• grok LOW — the ledger fixtures asserted the new migration's row but not its effect, so a
  repair that wrote 038 without applying the ALTER TABLE would still pass. Fixed: the repair
  scenario now asserts pin_order_key exists.
• codex MED — mobile computes Move Up/Down neighbours from all reorder-capable shells while
  rendering pins after env/project/search/snooze filtering, so a hidden pin can become the actual
  neighbour. Grok examined the same behaviour and filed it as upstream's own design (mobile
  "all shells" move vs web visible-list drag), not a merge weld — the fork's filters just make it
  more visible. Left as upstream ships it; changing it here would be a product decision beyond
  the sync. Worth a follow-up PR if you want mobile to match web.

Verification

• Full recursive typecheck clean across 17 packages.
• 2,284 tests pass; the only failure is the pre-existing CodexTextGeneration launch-args one,
  which reproduces on unmerged fork/dev.

Co-authored by @patroza

opened by Patrick Roza in chat thread Discord · Discord · T3