deps: the radix focus-scope patch targets 1.1.16 with no upstream issue while npm is at 1.3.1 — test stock 1.3.1 with the regression test and delete the patch if green #13466

Activity

1. os-zhuang

   added

   dependencies

   finding

   domain:uiobjectui ui stream: fix lands on the published library or apps — objectui execution seatobjectui ui stream: fix lands on the published library or apps — objectui execution seat

   on Oct 24, 2026
2. objectstack-fleet

   added

   area:devpathThe road — create, dev, verify, publish/install, connect an agent, iterateThe road — create, dev, verify, publish/install, connect an agent, iterate

   pm:queue

   priority:p3

   and removed

   finding

   on Oct 24, 2026
3. objectstack-fleet commented on Oct 24, 2026

   

   objectstack-fleetBot

   on Oct 24, 2026

   Contributor

   More actions

   Triage: first grade — dependencies · priority:p3 · domain:ui · area:devpath · pm:queue · triage seat · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-24T17:06Z

   ⛔ Not a claim, ⛔ not a dispatch.
   finding comes off.

   • Follow the patch README's own recipe: run dialog-popover-focus-scope.test.tsx on stock 1.3.1. If it is green, delete the patch. If it is red, keep the patch and add the upstream issue link the README lacks.

   ---

   Generated by Claude Code
4. objectstack-fleet

   added

   pm:dispatched

   and removed

   pm:queue

   on Oct 25, 2026
5. objectstack-fleet

   assigned

   os-elon-musk

   on Oct 25, 2026
6. objectstack-fleet commented on Oct 25, 2026

   

   objectstack-fleetBot

   on Oct 25, 2026

   Contributor

   More actions

   Claim: PM loop round 1 · 2026-10-25T02:31Z
   Session: session_0196W5VLn6ZrtibyAZrR8oiK (the card session that works this card; claim written by the domain:ui seat 4 session session_01XWGTsmqrtrt1Zqh1ioWqsT)
   Account: os-elon-musk
   Branch: claude/project-thread-ugocmh (platform-assigned at spawn; the card session can push only this branch)
   Worktree: objectui-issue-13466
   Domain: domain:ui
   Seat: domain:ui#4
   File surface: patches/@radix-ui__react-focus-scope.patch, patches/README.md, the pnpm.patchedDependencies entry in the root package.json (that entry only), pnpm-lock.yaml as pnpm install regenerates it, the @radix-ui/react-focus-scope range wherever a manifest pins it; packages/components/src/__tests__/dialog-popover-focus-scope.test.tsx only if the bump needs it; .changeset/13466-*.md if a published package's dependencies change (stop on breach; explain in the report)
   Container & model: S, mode:thread, model: default tier (dispatch-gates --tier --repo objectstack-ai/objectui: no path-derived mandate), effort: medium
   Clause-②: no
   Metadata-protocol: none
   Responsibility: n/a — not a defect card
   Thread-read: 6387085430
   Premise: holds — on objectui main 23e912e patches/@radix-ui__react-focus-scope.patch exists, root package.json:214 carries patchedDependencies, and the regression test sits at packages/components/src/__tests__/dialog-popover-focus-scope.test.tsx
   Serial constraints cleared: seat 4's PR #13481 (in the merge queue) edits manifests' React peers and the lockfile — merge main before opening and again before landing; no other in-flight card names the patch
   Writes: all PR, comment and label writes go only through objectstack's relay (scripts/pm/fleet-write); never write with GitHub MCP.

   • Ruling section: filed under the maintainer's 「同意：建议新建：与合并无关、独立成卡的产品/契约项（可现在建）」; triage 6387085430: the patch README's own recipe — stock 1.3.1, patch removed, run the regression test.
   • Green: delete the patch, its patchedDependencies entry and patches/README.md's section (the whole file if nothing else is listed); move the dependency to the version tested. Also run the components package's full suite on the bump.
   • Red: keep the patch and change nothing in patches/; report the failing assertion. ⛔ Filing an upstream radix issue is an outside-the-project write: the dev does not do it; the seat brings it to the maintainer. In that case no PR is owed — hand back the reading.
   • Measured before this claim (card session, worktree off main 6f629a7): patched 1.1.16 → regression test 3/3; stock 1.3.1 → 1 failed / 2 passed (the race leg, line 95). RED, so the patch stays. The upstream filing waits on the maintainer; once a link exists, the PR adds it, a measured-on line and a real exit condition to patches/README.md.
7. hotlong commented on Oct 25, 2026

   

   hotlong

   on Oct 25, 2026

   Contributor

   More actions

   The upstream issue already exists, so no new filing is needed · cloud ops seat, for the maintainer, 2026-10-25.

   The card session read RED: stock 1.3.1 fails the race leg, 1 failed / 2 passed. The open item was the maintainer's upstream filing. I searched radix-ui/primitives before filing anything, and the same root cause is already open:

   • [FocusScope] A Suspense hide and reveal of an open Dialog drops its scope from the stack; the next modal cannot pause it (focus ping-pong, RangeError) radix-ui/primitives#4226 (open, 2026-10-19): "[FocusScope] A Suspense hide and reveal of an open Dialog drops its scope from the stack; the next modal cannot …". The cleanup's setTimeout schedules focusScopesStack.remove(scope), the effect re-runs and adds the scope again, and then the stale timer evicts the live scope. That is exactly what patches/@radix-ui__react-focus-scope.patch cancels. Only the trigger differs: there it is a Suspense hide/reveal, here a container ref flicker (the regression test's key swap).
   • Related, closed: [FocusScope] Unmount setTimeout is never cleared and uses realm globals, so it can throw after test environment teardown radix-ui/primitives#4148, about the same unmount setTimeout never being cleared (the realm-globals variant).

   For the card session: use radix-ui/primitives#4226 as the upstream link in patches/README.md. Add the measured-on line (stock 1.3.1 → race leg red, 2026-10-25) and the exit condition "#4226 fixed in a released @radix-ui/react-focus-scope and the regression test green on it with the patch removed". That is the PR the claim already describes for the red branch.

   Commenting our trigger on #4226 (a second reproduction) would be a write to the radix repo, so it waits on the maintainer's say-so and isn't done here.
8. objectstack-fleet commented on Oct 25, 2026

   

   objectstack-fleetBot

   on Oct 25, 2026

   Contributor

   More actions

   Landed · domain:ui#4 · 2026-10-25T14:00Z

   PR #13529 merged at 2026-10-25T13:55Z, merge commit e6a06d2d2eed7dc7ceff82dfcfd8da50258fdfec. A read of origin/main patches/README.md shows the radix-ui/primitives#4226 upstream link and the new Remove when line. The card took the red branch: stock 1.3.1 fails the race case, so the patch stays, and its exit condition now names an upstream item that can trigger it. The card closes completed and pm:dispatched comes off in this stroke. objectui merges publish nothing.

   • Left open, the maintainer's call: adding this repo's trigger (a container ref flicker) to fix(metadata-admin): DatasourcePreview reads only keys DatasourceSchema accepts (#4131) #4226 as a second reproduction would be a write to the radix repo, so it is not done.
   • CI aside: the post-merge workflow "Close issues referenced in other repositories" went red on the PR head (run 40199822807). It runs after the merge and doesn't gate it. It most likely tripped on the cross-repo radix links in the PR body; this seat did not read its log.

   ---

   Generated by Claude Code
9. objectstack-fleet

   removed

   pm:dispatched

   on Oct 25, 2026
10. objectstack-fleet

    unassigned

    os-elon-musk

    on Oct 25, 2026
11. objectstack-fleet

    closed this as completedon Oct 25, 2026