Skip to content

[$250] Scan - Navigating to thread via "Explain" link on failed scan expense, shows skeleton loader #94411

Closed

Description

@applause-bot

If you haven’t already, check out our contributing guidelines for onboarding and email contributors@expensify.com to request to join our Slack channel!


Version Number: 9.4.18-0
Reproducible in staging?: Yes
Reproducible in production?: No
If this was caught during regression testing, add the test name, ID and link from BrowserStack: #93532
Email or phone of affected tester (no customers): ibellicotest+2766@gmail.com
Issue reported by: Applause Internal Team
Bug source: Exploratory - Significant User Experience Deterioration
Device used: Motorola MotoG60 / Android 12 (Hybrid app) - Windows 11 / Chrome
App Component: Money Requests

Action Performed:

Prerequisite: Account has at least one workspace.

  1. Open the Expensify app.
  2. Tap on FAB > "Create Expense" > "Scan"
  3. Create a scan expense with a receipt that will trigger a failure and submit it to workspace chat.
  4. Open the just created expense.
  5. Wait until receiving a Concierge message showing the scanning failure.
  6. Tap on "Explain" at the end of that message,
  7. Note that you land on a thread, and that skeleton loader is displayed for some time before messages are visible.

Expected Result:

After tapping on "Explain" on Concierge message after scan failure, user should land on a new created thread that should automatically load.

Actual Result:

Skeleton loader is displayed for some time when user navigates to thread after tapping on "Explain" link on Concierge message after scan expense failure.

Workaround:

Unknown

Platforms:

  • Android: App
  • Android: mWeb Chrome
  • iOS: App
  • iOS: mWeb Safari
  • iOS: mWeb Chrome
  • Windows: Chrome
  • MacOS: Chrome Safari

Screenshots/Videos

Bug7189308_1782277258832.Explain.mp4

View all open jobs on GitHub

Applause Internal Information
Exported by: Julia Ponikarchuk
Bug ID: 7189308

Upwork Automation - Do Not Edit

Activity

  1. added
    DeployBlockerCashThis issue or pull request should block deployment
    BugSomething is broken. Auto assigns a BugZero manager.
    on Jun 24, 2026
  2. applause-bot commented on Jun 24, 2026

    @applause-bot
    Author

    While Applause doesn't embed attachments for this bug, here are the links for quick view:
    https://platform.applause.com/services/links/v1/external/46abd74222963c01bf5de3a5516f77b064016ad32a030448b85d1a98bc4312da

  3. melvin-bot commented on Jun 24, 2026

    @melvin-bot

    You have been assigned to this deploy blocker because you recently merged this PR: #93532

  4. melvin-bot commented on Jun 24, 2026

    @melvin-bot

    💬 A slack conversation has been started in #expensify-open-source

  5. github-actions commented on Jun 24, 2026

    @github-actions
    Contributor

    👋 Friendly reminder that deploy blockers are time-sensitive ⏱ issues! Check out the open `StagingDeployCash` deploy checklist to see the list of PRs included in this release, then work quickly to do one of the following:

    1. Identify the pull request that introduced this issue and revert it.
    2. Find someone who can quickly fix the issue.
    3. Fix the issue yourself.
  6. MelvinBot commented on Jun 24, 2026

    @MelvinBot
    Contributor

    Issue Analysis

    This is expected behavior, not a regression from #93532 — recommend removing the DeployBlockerCash label and not reverting.

    The skeleton loader is generic to every Explain thread, not specific to failed scans. Tapping any inline Explain link runs the shared explain() flow: if the explanation thread doesn't already exist, it creates an optimistic Concierge thread and fires OpenReport, and the thread screen shows ReportActionsSkeletonView until that request returns. PR #93532 added no scan-specific load logic — it only wired RECEIPTSCANFAILED messages into the same ReportActionItemMessageWithExplain → explain() → createChildReport() → openReport() path already used by MovedTransactionAction, ExportIntegration, ModifiedExpenseContent, UnreportedTransactionAction, and ApprovalFlowContent.

    Two facts back this up:

    • Not reproducible in production (per the report) — a code-path bug would reproduce everywhere. The staging-only nature points to OpenReport round-trip latency / the explanation thread not yet existing, i.e. an environmental/timing difference, not a code change.
    • The same skeleton already appeared pre-PR when opening this same Explain thread via the right-click / long-press context menu (which gated on hasReasoning() before this PR). [Payment due @brunovjk] Show Explain button on receipt scan failed system messages #93532 only made the affordance more discoverable.

    Recommendation: Remove DeployBlockerCash. If the brief skeleton on freshly-created Explain threads is considered a real UX problem, it should be a separate, non-blocking issue scoped to the generic Explain/OpenReport loading experience — it affects all Explain actions equally and predates this PR.

    Investigation details (evidence)

    Shared entry point — every Explain link runs the same handler:

    Thread creation / fetch — the skeleton source:

    • Report/index.ts:2552-2592 — const report = childReport ?? createChildReport(...). If childReport exists, it navigates straight to a populated thread (no skeleton). If not, a new optimistic thread is created.
    • Report/index.ts:2485-2543 — createChildReport calls openReport({ isNewThread: true, ... }) for a brand-new thread.
    • Report/index.ts:1500-1552 — openReport optimisticData sets isLoadingInitialReportActions: true; successData clears it. The skeleton-visible duration == the OpenReport round-trip latency.

    Skeleton gate — generic, no scan-failed branch:

    • ReportActionsView.tsx:178 and :218-222 — shouldShowSkeletonForInitialLoad keyed on isLoadingInitialReportActions; renders ReportActionsSkeletonView. No receipt-scan logic here.

    Identical wiring across all Explain consumers — childReport derived from the same Onyx key (REPORT + action.childReportID) and the same originalReport from ActionContentRouter:

    • ReceiptScanFailedContent.tsx:39, 48-55 vs. MovedTransactionAction.tsx / UnreportedTransactionAction.tsx / ExportIntegration.tsx (byte-for-byte the same childReport derivation). The scan-failed-only props (reportID, parentReportID, etc.) are used solely for the canEdit IOU-actor lookup, never for navigation.

    Alternative explanation considered: a genuine new slow path introduced by the PR. Ruled out — the PR's diff (ReceiptScanFailedContent.tsx, ActionContentRouter.tsx, OriginalMessage.ts, tests) touches only rendering/typing, not the explain()/openReport machinery.

    Optional confirmation (not required): compare OpenReport latency (with isNewThread=true, originating from Explain taps) on staging vs production to confirm the staging-only repro is latency/thread-freshness, not a code difference.


    Next Steps for Contributor+ team:
    To accept this assessment: remove the DeployBlockerCash label (no revert needed). Optionally file a separate non-blocking issue for the generic Explain-thread loading UX.
    To refine: @MelvinBot <your feedback>
    To reject: explain why you believe this is a true regression from #93532.

  7. added and removed
    DeployBlockerCashThis issue or pull request should block deployment
    on Jun 24, 2026
  8. brunovjk commented on Jun 24, 2026

    @brunovjk
    Contributor

    I don’t think this is a regression from #93532. The PR only exposed the existing Explain flow inline for RECEIPT_SCAN_FAILED.

    The skeleton is the normal shared loading state while a new Explain child thread is opened, so this looks like generic Explain-thread behavior rather than something introduced by the PR.

  9. 29 remaining items

  10. brunovjk commented on Jul 6, 2026

    @brunovjk
    Contributor

    looks like to be not related to #93532 ... should i review the proposals ? @brunovjk @mollfpr

    I think so, sorry.

  11. melvin-bot commented on Jul 9, 2026

    @melvin-bot

    @mollfpr, @abzokhattab Whoops! This issue is 2 days overdue. Let's get this updated quick!

  12. melvin-bot commented on Jul 10, 2026

    @melvin-bot

    @mollfpr, @abzokhattab Huh... This is 4 days overdue. Who can take care of this?

  13. mvtglobally commented on Jul 11, 2026

    @mvtglobally

    Issue not reproducible during KI retests. (First week)

  14. abzokhattab commented on Jul 12, 2026

    @abzokhattab
    Contributor

    issue is not reproducible for me as well ... lets close this issue @mollfpr

    Screen.Recording.2026-07-12.at.17.55.11.mov
  15. melvin-bot commented on Jul 13, 2026

    @melvin-bot

    📣 It's been a week! Do we have any satisfactory proposals yet? Do we need to adjust the bounty for this issue? 💸

  16. melvin-bot commented on Jul 15, 2026

    @melvin-bot

    @mollfpr, @abzokhattab Whoops! This issue is 2 days overdue. Let's get this updated quick!

  17. mvtglobally commented on Jul 17, 2026

    @mvtglobally

    Issue not reproducible during KI retests. (Second week)

  18. brunovjk commented on Jul 17, 2026

    @brunovjk
    Contributor

    @mollfpr, can you help me close this issue? Thanks.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Labels

BugSomething is broken. Auto assigns a BugZero manager.DailyKSv2EngineeringExternalAdded to denote the issue can be worked on by a contributorHelp WantedApply this label when an issue is open to proposals by contributors

Type

No type

Projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions