[Onyx audit] Migrate keys to RAM-only - part 1/2

[JKobrynski]

Explanation of Change

This PR is part 1/2 of RAM-only migration - it migrates the obvious keys.

Fixed Issues

$ #80091
PROPOSAL: N/A

Tests

1. App Launch & Deep Linking (IS_CHECKING_PUBLIC_ROOM)

1. Launch the app and confirm the splash screen disappears and you land on the home screen without hanging
2. Open a deep link to a public room while logged out (e.g. a shared report link) — confirm the app correctly navigates to the public room or shows the sign-in page
3. Open a deep link to a public room while logged in — confirm you are taken directly to the report
4. Open a deep link to a private room while logged out — confirm you are redirected to the sign-in page
5. Kill and restart the app, then repeat steps 1–4 to verify no stale state persists across restarts

---

2. App Update Modals (UPDATE_AVAILABLE, UPDATE_REQUIRED)

It's difficult to test these, as the related actions are only triggered by API

6. Make sure no "Update required"/"Update available" modal is shown by mistake when launching the app

---

3. Search / User Selection Loading States (IS_SEARCHING_FOR_REPORTS)

New Chat

7. Go to New Chat (+ button > New Chat), type a search term in the search field — confirm a loading indicator appears while results are being fetched from the server and disappears once results load

Money Request (Expense) Participant Selector

8. Start creating a new expense (Request Money / Split Expense), and in the participant selector search for a user — confirm the loading indicator appears during server search
9. Confirm the results load correctly and you can select a participant

Workspace Invite

10. Go to a Workspace > Members > Invite, search for a user to invite — confirm the loading spinner appears during the search and results load

Search Router

11. Open the global Search (magnifying glass icon), type a query — confirm loading indicator appears and search results populate

---

4. Wallet / Onfido Identity Verification (WALLET_ONFIDO)

12. Go to Settings > Wallet > Enable Payments and start the wallet activation flow
13. When you reach the Onfido step, confirm the privacy policy screen appears first (not stale data from a previous attempt)
14. Accept the privacy policy — confirm a loading indicator shows while the Onfido SDK token is being fetched
15. If identity verification fails, confirm the error is displayed and you are sent back to the privacy policy screen on retry
16. Navigate away from the Onfido flow and then navigate back — confirm the flow restarts cleanly (the privacy policy screen shows again, not stale SDK data)
17. Kill and restart the app, then go back to the Onfido flow — confirm no stale wallet/onfido data persists

---

5. General Regression Checks

18. Confirm the app works correctly offline — toggle airplane mode, navigate around, and confirm no crashes related to these keys
19. Confirm logging out and logging back in works without issues (no stale RAM-only key state leaking across sessions)
20. Confirm there are no console errors or warnings related to Onyx keys on app startup

• Verify that no errors appear in the JS console

Offline tests

N/A

QA Steps

Same as Tests section above

• Verify that no errors appear in the JS console

PR Author Checklist

• I linked the correct issue in the ### Fixed Issues section above
• I wrote clear testing steps that cover the changes made in this PR
  • I added steps for local testing in the Tests section
  • I added steps for the expected offline behavior in the Offline steps section
  • I added steps for Staging and/or Production testing in the QA steps section
  • I added steps to cover failure scenarios (i.e. verify an input displays the correct error message if the entered data is not correct)
  • I turned off my network connection and tested it while offline to ensure it matches the expected behavior (i.e. verify the default avatar icon is displayed if app is offline)
  • I tested this PR with a High Traffic account against the staging or production API to ensure there are no regressions (e.g. long loading states that impact usability).
• I included screenshots or videos for tests on all platforms
• I ran the tests on all platforms & verified they passed on:
  • Android: Native
  • Android: mWeb Chrome
  • iOS: Native
  • iOS: mWeb Safari
  • MacOS: Chrome / Safari
• I verified there are no console errors (if there's a console error not related to the PR, report it or open an issue for it to be fixed)
• I verified there are no new alerts related to the canBeMissing param for useOnyx
• I followed proper code patterns (see Reviewing the code)
  • I verified that any callback methods that were added or modified are named for what the method does and never what callback they handle (i.e. toggleReport and not onIconClick)
  • I verified that comments were added to code that is not self explanatory
  • I verified that any new or modified comments were clear, correct English, and explained "why" the code was doing something instead of only explaining "what" the code was doing.
  • I verified any copy / text shown in the product is localized by adding it to src/languages/* files and using the translation method
    • If any non-english text was added/modified, I used JaimeGPT to get English > Spanish translation. I then posted it in #expensify-open-source and it was approved by an internal Expensify engineer. Link to Slack message:
  • I verified all numbers, amounts, dates and phone numbers shown in the product are using the localization methods
  • I verified any copy / text that was added to the app is grammatically correct in English. It adheres to proper capitalization guidelines (note: only the first word of header/labels should be capitalized), and is either coming verbatim from figma or has been approved by marketing (in order to get marketing approval, ask the Bug Zero team member to add the Waiting for copy label to the issue)
  • I verified proper file naming conventions were followed for any new files or renamed files. All non-platform specific files are named after what they export and are not named "index.js". All platform-specific files are named for the platform the code supports as outlined in the README.
  • I verified the JSDocs style guidelines (in STYLE.md) were followed
• If a new code pattern is added I verified it was agreed to be used by multiple Expensify engineers
• I followed the guidelines as stated in the Review Guidelines
• I tested other components that can be impacted by my changes (i.e. if the PR modifies a shared library or component like Avatar, I verified the components using Avatar are working as expected)
• I verified all code is DRY (the PR doesn't include any logic written more than once, with the exception of tests)
• I verified any variables that can be defined as constants (ie. in CONST.ts or at the top of the file that uses the constant) are defined as such
• I verified that if a function's arguments changed that all usages have also been updated correctly
• If any new file was added I verified that:
  • The file has a description of what it does and/or why is needed at the top of the file if the code is not self explanatory
• If a new CSS style is added I verified that:
  • A similar style doesn't already exist
  • The style can't be created with an existing StyleUtils function (i.e. StyleUtils.getBackgroundAndBorderStyle(theme.componentBG))
• If new assets were added or existing ones were modified, I verified that:
  • The assets are optimized and compressed (for SVG files, run npm run compress-svg)
  • The assets load correctly across all supported platforms.
• If the PR modifies code that runs when editing or sending messages, I tested and verified there is no unexpected behavior for all supported markdown - URLs, single line code, code blocks, quotes, headings, bold, strikethrough, and italic.
• If the PR modifies a generic component, I tested and verified that those changes do not break usages of that component in the rest of the App (i.e. if a shared library or component like Avatar is modified, I verified that Avatar is working as expected in all cases)
• If the PR modifies a component related to any of the existing Storybook stories, I tested and verified all stories for that component are still working as expected.
• If the PR modifies a component or page that can be accessed by a direct deeplink, I verified that the code functions as expected when the deeplink is used - from a logged in and logged out account.
• If the PR modifies the UI (e.g. new buttons, new UI components, changing the padding/spacing/sizing, moving components, etc) or modifies the form input styles:
  • I verified that all the inputs inside a form are aligned with each other.
  • I added Design label and/or tagged @Expensify/design so the design team can review the changes.
• If a new page is added, I verified it's using the ScrollView component to make it scrollable when more elements are added to the page.
• I added unit tests for any new feature or bug fix in this PR to help automatically prevent regressions in this user flow.
• If the main branch was merged into this PR after a review, I tested again and verified the outcome was still expected according to the Test steps.

Screenshots/Videos

Android: Native

Android: mWeb Chrome

iOS: Native

iOS: mWeb Safari

MacOS: Chrome / Safari

web-compressed.mov

[codecov]

Codecov Report

✅ Changes either increased or maintained existing code coverage, great job!

Files with missing lines	Coverage Δ
src/DeepLinkHandler.tsx	80.55% <100.00%> (+3.13%)	⬆️
src/Expensify.tsx	88.79% <100.00%> (ø)
src/ONYXKEYS.ts	100.00% <ø> (ø)
...ponents/BaseVacationDelegateSelectionComponent.tsx	0.00% <ø> (ø)
...ponents/Search/FilterDropdowns/UserSelectPopup.tsx	2.63% <ø> (+0.06%)	⬆️
...c/components/Search/SearchFiltersChatsSelector.tsx	0.00% <ø> (ø)
...rc/components/Search/SearchRouter/SearchRouter.tsx	0.00% <ø> (ø)
src/libs/actions/App.ts	47.30% <ø> (ø)
src/libs/actions/QueuedOnyxUpdates.ts	100.00% <ø> (ø)
src/pages/EnablePayments/OnfidoStep.tsx	0.00% <ø> (ø)
... and 22 more
... and 9 files with indirect coverage changes

[melvin-bot]

@Krishna2323 @roryabraham One of you needs to copy/paste the Reviewer Checklist from here into a new comment on this PR and complete it. If you have the K2 extension, you can simply click: [this button]

[Krishna2323]

I’ll be reviewing this shortly.

[Krishna2323]

@JKobrynski Yes, I was testing with the latest changes and had also verified that the bug was not happening on the main branch. I was testing by navigating to a public room created from a different account. Could you please try with new-expensify://r/7410719245147364, and before closing the app, navigate back to the home screen?

[JKobrynski]

@Krishna2323 sure thing, I will test this exact scenario

[JKobrynski]

@Krishna2323 I just tested with the report that you mentioned above and it seems to be working fine

android.mov

Have you checked again on your end? 😄

[Krishna2323]

Will test again today.

[JKobrynski]

It is possible that there's been a pre-existing race condition in DeepLinkHandler.tsx that is just more visible with the changes introduced in this PR. If that's the case, it would explain why it works for me and it doesn't work for you. I'm currently investigating the related code and potential fixes. I think it'd be best to retest just to make sure that we were testing the same version (as this PR is updated pretty much daily) and then test again after the fix is introduced - that's assuming I will be able to find an easy fix to the race condition. Will keep you posted, thanks for doing so much testing 🙏

[Krishna2323]

@JKobrynski It still doesn't work for me. One thing I forgot to tell you is that I'm running the standalone app.

Monosnap.screencast.2026-03-09.18-30-02.mp4

[JKobrynski]

@Krishna2323 I'm using standalone too, so all the same.

Ok so looking at this it seems like it must be the race condition. I'm working on a fix, I'll let you know when it's ready so you can test again.

[JKobrynski]

@Krishna2323 ok I've implemented a fix, I tested this flow and some other ones that are also related, could you please test the latest version again and let me know if it helped? 🙏

[Krishna2323]

@JKobrynski I can still repro the issue, I'm trying to investigate it now...

Monosnap.screencast.2026-03-10.04-23-53.mp4

[Krishna2323]

@JKobrynski seems like this is the cause, I'm trying to apply a fix locally to confirm.

Line 319 (very first render):

isCheckingPublicRoom: false meta: {"status":"loaded"}

Line 324-325 (effect fires immediately):

isCheckingPublicRoom effect fired, value: false
setAttemptedToOpenPublicRoom(true)

Line 362 (NavigationRoot renders too early):

render: hasAttemptedToOpenPublicRoom: true initialUrl: null

Line 462 (deep link resolves much later - TOO LATE):

getInitialURL resolved: new-expensify://r/7410719245147364

The root cause: On the very first render, isCheckingPublicRoom is already false (not true). This means the old stored value false from before the RAM-only migration is being loaded from storage despite the key being configured as RAM-only. The initialKeyStates: true we added is being overridden by the stale stored value.

This causes a cascade:

1. isCheckingPublicRoom = false on first render (stale stored value)
2. setAttemptedToOpenPublicRoom(true) fires immediately
3. NavigationRoot renders with initialUrl: null
4. Linking.getInitialURL() resolves ~400ms later with the deep link URL
5. initialUrl updates, but React Navigation ignores prop changes to initialUrl after mount

This is the exact scenario the Onyx team discussed in the issue -- JKobrynski warned that IS_CHECKING_PUBLIC_ROOM would "read the false value from storage on each init, and never default to true." The fix was supposed to be Onyx 3.0.41 (PR #740 - prevent storage reads for RAM-only keys), but it's clearly not working for this key.

This is a bug you should report to the PR author. The IS_CHECKING_PUBLIC_ROOM key cannot safely be migrated to RAM-only until the Onyx library correctly prevents loading stale stored values for keys that were previously regular keys.

[Krishna2323]

@JKobrynski please check this, it fixed the navigation issue:

What was happening:

NavigationRoot computes its navigation state only once on first render (via useMemo(() => ..., [])). For authenticated users, openReportFromDeepLink doesn't navigate programmatically -- it relies on NavigationRoot receiving the correct initialUrl on its very first mount.

The problem: DeepLinkHandler calls doneCheckingPublicRoom() before Linking.getInitialURL() resolves (which is async). This causes NavigationRoot to mount with initialUrl = null. When the actual deep link URL arrives moments later, it's too late -- NavigationRoot already initialized without it.

How it was fixed (2 changes):

1. Expensify.tsx: Changed initialUrl state from useState(null) to useState(undefined), and added initialUrl !== undefined to the NavigationRoot rendering gate. This prevents NavigationRoot from mounting until the deep link URL has actually been resolved.

2. DeepLinkHandler.tsx: Added an initialUrlProcessed ref so the isAuthenticated safety-net effect doesn't call doneCheckingPublicRoom() before getInitialURL() has resolved.

Monosnap.screencast.2026-03-10.05-27-06.mp4

[JKobrynski]

@Krishna2323 first of all thank you so much for doing all that on your end as I wasn't able to reproduce the issue 🙏 I'm reaching out to you on Slack with some questions!

[JKobrynski]

@Krishna2323 I've applied the following improvements to the fix that you've suggested (and the one I implemented just before that):

• Add .catch() and a timeout to getInitialURL() - handle scenarios where the method fails for some reason or doesn't resolve in <10seconds (this can be changed as necessary)
• Add a stale closure guard in the effect cleanup - prevents duplicate openReportFromDeepLink() calls when dependencies change
• Document the safety-net's behavior more explicitly

I'm going to share more details about my investigation in the issue but I think we can test this PR again.

[mountiny]

Nice one, great job! @Krishna2323 thanks for helping

[Krishna2323]

@codex review

[Krishna2323]

Works well 🎉

fix_part_1.mp4

fix_part_2.mp4

[Krishna2323]

@JKobrynski Is there anything left? I tried checking for similar cases but found that it shouldn’t happen anywhere else.

Here's what I found. Beyond the NavigationRoot issue we already fixed, there are a few other places with similar one-time-computation patterns, but none of them have the same cold-start severity:

---

Already fixed (the critical one):

• NavigationRoot.tsx lines 104-142 -- useMemo(() => ..., []) reads initialUrl once. This is the bug we fixed by gating NavigationRoot mounting on initialUrl !== undefined.

Other patterns found but low/no risk:

1. useInitialValue hook (src/hooks/useInitialValue.ts) -- Used in TimezoneSelectPage.tsx to compute timezone list once. It reads timezone.selected from personal details on first render. If personal details load late, isSelected could be wrong. But this is a settings page (not cold start), and the user navigates there manually -- by that time Onyx data is always loaded. No risk.

2. WorkspaceOwnerRestrictedAction (index.native.tsx line 22) -- useMemo(() => Navigation.getActiveRoute(), []) freezes the active route. But this component only renders when a user navigates to it (never on cold start), so the navigation state is always ready. No risk.

3. MapView components -- These have initialState in their useMemo deps (not empty), so they recompute when props change. No risk.

4. SignInModal.tsx -- useMemo(() => isMobileSafari(), []) -- purely synchronous platform check. No risk.

5. InitialURLContextProvider -- Calls Linking.getInitialURL() but uses useState/setInitialURL (not useMemo), so it updates correctly when the promise resolves. No risk.

6. SplitRouter.ts -- getInitialState reads preserved navigator state. It already has a guard checking navigationRef.isReady(). Low risk.

Bottom line: The cold-start race condition was unique to the NavigationRoot + DeepLinkHandler + Expensify interaction. No other code in the codebase has this same pattern of gating a component that reads an async value only once on mount during cold start. The author's fix is sufficient.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4e9b00ca33

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[chatgpt-codex-connector]

Handle late getInitialURL resolution after timeout

Racing Linking.getInitialURL() against a 10s null timeout unblocks startup, but it also permanently drops deep links when the native bridge returns the initial URL after that timeout (a case this comment already calls out in HybridApp). Once the timeout wins, onInitialUrl(null) runs and the eventual URL resolution is ignored, so users can land on the default route instead of the intended deep-linked report on slower/lagging startup paths.

Useful? React with 👍 / 👎.

[JKobrynski]

@Krishna2323 do you have any thoughts on this ☝️ ?

[JKobrynski]

Is there anything left? I tried checking for similar cases but found that it shouldn’t happen anywhere else.

These are just some suggestions for further investigation on similar cases, and potential improvements to consider!

[Krishna2323]

I'm going to share more details about my investigation in the issue but I think we can test this PR again.

@JKobrynski I was asking about this :)

[JKobrynski]

@Krishna2323 posted on the issue yesterday!