Answer page: "Report a problem" opens as a sheet, not a panel behind the composer - #2541
Conversation
…flow panel The list of problem types opened as a disclosure below the utility row, and on a phone it opened partly behind the fixed composer. It could not scroll itself clear: every scripted scroll that would do it is a DOWNWARD scroll, a downward scroll is what hides the phone chrome, and closing the panel then shrank the page back to the top without generating the upward travel that reveals the chrome again — so the composer stayed gone. `ui-smoke`'s critical answer journey caught exactly that, twice, and the auto-scroll was reverted both times. Reachability was never the defect. Measured at 390x844, the last option cleared the composer by 180px at full scroll. The defect was that the list LOOKED complete when it was not. A Sheet answers both halves at once. It owns its own scrollport above the composer, so nothing is clipped and no page scroll is needed, and it is the same overlay the safety-findings control beside it already opens — `mobilePlacement` defaults to "bottom", so it rises from the bottom on a phone and is a centred dialog from `sm:` up. The trigger becomes a dialog opener (`aria-haspopup="dialog"`, focus returned to it on close) rather than a disclosure. `AnswerFeedbackPanel` gains `chrome="bare"`, which drops its card border and its own heading pair for a host that already carries them: the sheet's title and description ARE `answerFeedbackQuestion`, now exported so the two cannot drift apart, and asking the same question twice in one overlay is the duplication that prop exists to avoid. The default `chrome="card"` leaves the full-tone panel in `visual-evidence.tsx` untouched. Verification: `npm run test` 12055 passed | 1 skipped; `npm run lint` and `npm run typecheck` exit 0; the six affected Chromium journeys pass, including the critical answer flow that the reverted auto-scroll broke. A new `ui-smoke` case pins the phone geometry the old panel failed — at 390x844 every option must sit inside the viewport with the sheet at rest, at the repo's 48px tap floor, and Escape must return focus to the trigger and leave the composer where it was. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01NkKHznd5E5KLnKbowRTLd8
|
Important
This repository does not receive automatic reviews because it has fewer than 10 stars. ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Team Run ID: Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
This pull request has been ignored for the connected project Preview Branches by Supabase. |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_047be90b-02b8-4318-b886-71e53f83ddba) |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: c7d5536105
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
@codex resolve actionable Codex review findings for this pull request and current head using the repository instructions. This is the pull request's single automatic repair pass: do not perform a fresh review, create new standalone findings, or request another review. Work only the existing unresolved Codex threads on the current head. The only repository destination is BigSimmo/Database, and the only branch destination is the pull request head branch claude/answer-page-handover-c2qlwy at starting commit c7d5536; never publish fixes to a detached or synthetic work branch and never create a stacked pull request. Use the authenticated GitHub connector to commit each approved fix to BigSimmo/Database:claude/answer-page-handover-c2qlwy, then verify that the pull request head contains the published commit before reporting success. Always fix P0 and P1 findings. For P2 and lower findings, fix only clear, scoped, low-risk issues; otherwise disposition them with a concise reason. For a fixed thread, reply with as the first line and as the second line. For a no-code disposition, use followed by . These result markers authorize the workflow to close that exact thread only after it verifies a fixed commit is the pull request head; a local-only commit is not a fix. If publication or verification fails, do not use either result marker, do not claim success, and leave the thread open with the blocker. If human input or new authorization is required, do the same. Finish only after every actionable thread is fixed or dispositioned and closed, or explicitly left open for a human decision. Do not update the branch from main, address unrelated reviews, broaden scope, or create more than one scoped fix commit. Do not use external APIs, paid services, credentials, dependency changes, or broad refactors unless explicitly authorized. Add targeted tests where behavior changes and run the narrowest relevant validation. |
Summary
Testing
|
…is visible Codex P2 on #2541, confirmed against the code. `submitAnswerFeedback` in ClinicalDashboard reports EVERY outcome through the page-level `actionNotice` alone — success, network failure, an expired feedback token, and synthetic demo answers — and that notice renders outside the portaled sheet. Leaving the sheet open put it behind the backdrop with the page under it inert, so choosing a problem type read as a tap that did nothing. The demo and expired-token paths are the worst of it: both return before `pendingFeedback` is ever set, so there was not even a spinner to explain the silence. Demo mode is how the app runs without Supabase configured, so that was the common case, not an edge one. This did not exist while the list was an in-flow disclosure — the notice was simply visible above it — so it is a regression introduced by the sheet in the preceding commit, not a pre-existing fault. Submitting now closes the sheet, which restores the notice and the focus return to the trigger. The affirmative thumb-up path was never in a sheet and is untouched. Verification: the new DOM case fails without the fix (`1 failed | 8 passed`) and passes with it, so it catches the regression rather than describing it; `npm run test` 947 files, 12058 passed | 1 skipped; `npm run lint` and `npm run typecheck` exit 0; focused browser proof on the owning spec — `tests/ui-smoke.spec.ts`, 106 passed (4.5m) — with the full Chromium suite left to CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01NkKHznd5E5KLnKbowRTLd8
…wy' into claude/answer-page-handover-c2qlwy
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_16d4eed9-8803-4e6c-9c21-f70e535a15cd) |
2 participants
Summary
Sheetrather than an in-flow disclosure below the utility row. On a phone the disclosure opened partly behind the fixed composer, and it could not scroll itself clear: every scripted scroll that would do it is a downward scroll, a downward scroll is what hides the phone chrome, and closing the panel then shrank the page back to the top without generating the upward travel that reveals the chrome again — so the composer stayed gone.ui-smoke's critical answer journey caught exactly that, twice, and the auto-scroll was reverted both times.mobilePlacementdefaults to"bottom", so it rises from the bottom on a phone and is a centred dialog fromsm:up.aria-haspopup="dialog", focus returned to it on close — rather than a disclosure.aria-expandedand thearia-controlstarget are kept.AnswerFeedbackPanelgainschrome="bare", which drops its card border and its own heading pair for a host that already carries them: the sheet's title and description areanswerFeedbackQuestion, now exported so the two cannot drift apart. Asking the same question twice in one overlay is the duplication that prop exists to avoid. The defaultchrome="card"leaves the full-tone panel invisual-evidence.tsxuntouched.Verification
npm run verify:ui—646 passed (25.0m), exit 0. This includes the critical answer journey that the reverted auto-scroll broke.npm run test(full offline unit suite) —Test Files 947 passed (947),Tests 12055 passed | 1 skipped (12056).npm run lint— exit 0 ([gate-receipts] recorded a pass for "lint:internal" (5999 input files)).npm run typecheck— exit 0.Verification not run:
npm run verify:pr-local— its heavy scope is lint + typecheck + the full unit suite + build-backed gates, each run directly on this tree and reported above with its own output.npm run verify:release,eval:retrieval:quality,eval:rag,eval:qualityandcheck:production-readinessare provider-backed and were not authorized; no retrieval, ranking, generation, ingestion or environment behaviour changed here.After the gate the branch merged
origin/main(documentation splits, a review record, an issue record, and the Ward Flow pinned-clock fix) with no conflict and no change to any file this branch touches.Risk and rollout
Clinical Governance Preflight
Clinical KB Database(sjrfecxgysukkwxsowpy)On the sixth item: nothing about source metadata or review status is read or rendered differently here. On the seventh: no decision-support behaviour changed — no claim, ranking, citation or verification output differs — so the classification is unaffected.
Notes
ui-smokecase pins the phone geometry the old panel failed: at 390×844, with the sheet at rest and no page scroll, every option must sit inside the viewport at the repo's 48px production tap floor (not the generic 44px, which reintroduces a known flake), the affirmative "Verified" option must be absent from a list opened to report a fault, and Escape must both return focus to the trigger and leave the composer where it was.🤖 Generated with Claude Code
https://claude.ai/code/session_01NkKHznd5E5KLnKbowRTLd8
Generated by Claude Code
Note
Low Risk
Presentation-only change to one answer feedback control; feedback options and submission behavior are unchanged aside from closing the modal so notices remain visible.
Overview
Report a problem on the answer utilities row now opens a
Sheetinstead of an in-flow panel under the thumbs-down control. That fixes mobile UX where the old list looked complete while options sat behind the fixed composer, and avoids the scroll trap that hid the phone chrome when trying to scroll the list into view.The thumbs-down trigger is wired as a dialog opener (
aria-haspopup="dialog", focus return viareturnFocusRef).AnswerFeedbackPaneladdschrome="bare"and sharedanswerFeedbackQuestioncopy so the sheet header owns the title/description without duplicating the card chrome. Choosing a problem type closes the sheet immediately so page-level outcome notices (success, errors, demo paths) are not hidden behind the modal backdrop.Tests and
ui-smokenow assert the portaled sheet, Escape dismissal, bare panel chrome, close-on-submit, and at 390×844 that every option is fully in the viewport at the 48px tap floor.Reviewed by Cursor Bugbot for commit ea9887e. Configure here.