[$250] Remove time-sensitive task and RBR for unresolved broken cards after 90 days

[melvin-bot]

Feature Description

Coming from here, when a personal or company card connection is broken and remains unresolved after 90 days, we want to:

• Remove the time-sensitive home task. See below what this looks like now.
• Remove the RBR leading to the broken connection error (but keep the error itself, so users can still fix it if they choose). This means:
  • Personal cards: Remove the indicator on Account and red dot on Wallet row in the account left hand bar. But keep the error on the card itself on the Wallet page so the card can still be fixed.
  • Company cards: Remove the red dots leading to the broken connection error on the company cards page in the workspace. But leave the error itself there so the user can still fix it.

This allows users with long-standing broken connections to continue using the app without active prompting, while still maintaining the option to fix the connection if desired.

Example of time-sensitive home task we'd remove after 90 days

Upwork Automation - Do Not Edit

• Upwork Job URL: https://www.upwork.com/jobs/~022057928818599135150
• Upwork Job ID: 2057928818599135150
• Last Price Increase: 2026-05-22
• Automatic offers:
• wildan-m | Contributor | 111659805

[melvin-bot]

Triggered auto assignment to Design team member for new feature review - @dubielzyk-expensify (NewFeature)

[joekaufmanexpensify]

@dubielzyk-expensify no mocks needed here, so removing you

[melvin-bot]

Triggered auto assignment to Contributor-plus team member for initial proposal review - @jayeshmangwani (External)

[wildan-m]

Proposal

Please re-state the problem that we are trying to solve in this issue.

When a personal or company card connection has been broken and unresolved for 90 days, we want to stop actively prompting the user: remove the time-sensitive home task and remove the RBR (red dot) that points at the broken connection — while keeping the underlying error so the user can still fix the connection if they choose.

What is the root cause of that problem?

There is no time component anywhere in the broken-card surfacing logic today, so a broken connection prompts the user indefinitely no matter how long it has gone unresolved.

Both the time-sensitive home task and the RBR are driven by the same derived Onyx value computed in cardFeedErrors.ts. For every card it calls isCardConnectionBroken(card), and when that is true it does two things: it adds the card to cardsWithBrokenFeedConnection / personalCardsWithBrokenConnection (the home "Time sensitive" section renders these as the Fix-connection task), and it sets isFeedConnectionBroken / shouldShowRBR (which drives the LHN account red dot, the workspace red dot, and the company-card feed banner).

App/src/libs/actions/OnyxDerived/configs/cardFeedErrors.ts

Lines 64 to 78 in 826c670

	const isFeedConnectionBroken = isCardConnectionBroken(card);
	// Track personal cards with broken feed connection
	if (isFeedConnectionBroken) {
	personalCardsWithBrokenConnection[card.cardID] = card;
	}
	const newFeedState: Omit<CardFeedErrorState, 'shouldShowRBR'> = {
	isFeedConnectionBroken,
	hasFeedErrors: !isEmptyObject(cardErrors),
	hasWorkspaceErrors: false,
	};
	const shouldShowRBR = getShouldShowRBR(newFeedState);
	personalCardStates.isFeedConnectionBroken ||= newFeedState.isFeedConnectionBroken;
	personalCardStates.hasFeedErrors ||= newFeedState.hasFeedErrors;
	personalCardStates.shouldShowRBR ||= shouldShowRBR;

App/src/libs/actions/OnyxDerived/configs/cardFeedErrors.ts

Lines 119 to 140 in 826c670

	const isFeedConnectionBroken = isCardConnectionBroken(card);
	const newFeedState: Omit<CardFeedErrorState, 'shouldShowRBR'> = {
	isFeedConnectionBroken: isFeedConnectionBroken || previousFeedErrors.isFeedConnectionBroken,
	hasFeedErrors: hasFeedErrors || previousFeedErrors.hasFeedErrors,
	hasWorkspaceErrors: hasWorkspaceErrors || previousFeedErrors.hasWorkspaceErrors,
	};
	const shouldShowRBR = getShouldShowRBR(newFeedState) || previousFeedErrors.shouldShowRBR;
	cardFeedErrors[feedNameWithDomainID] = {
	...newFeedState,
	shouldShowRBR,
	cardErrors,
	feedErrors,
	workspaceErrors,
	};
	// Track cards with broken feed connection
	if (isFeedConnectionBroken) {
	cardsWithBrokenFeedConnection[card.cardID] = card;
	}

isCardConnectionBroken itself only checks the latest scrape result code and has no notion of how long the connection has been failing:

App/src/libs/CardUtils.ts

Lines 1243 to 1248 in 826c670

	function isCardConnectionBroken(card: Card): boolean {
	if (card.pendingFields?.lastScrape) {
	return false;
	}
	return !!card.lastScrapeResult && !CONST.COMPANY_CARDS.BROKEN_CONNECTION_IGNORED_STATUSES.includes(card.lastScrapeResult);
	}

The home task consumes those same broken-card lists:

App/src/pages/home/TimeSensitiveSection/index.tsx

Lines 113 to 147 in 826c670

	// Get company cards with broken connections (for admins)
	const brokenCompanyCardConnections: BrokenCompanyCardConnection[] = [];
	const cardsWithBrokenConnection = cardFeedErrors.cardsWithBrokenFeedConnection;
	if (cardsWithBrokenConnection && adminPolicies) {
	for (const card of Object.values(cardsWithBrokenConnection)) {
	if (!card?.fundID) {
	continue;
	}
	// Find the policy associated with this card's fundID (workspaceAccountID)
	const cardFundID = Number(card.fundID);
	const matchingPolicy = adminPolicies.find((policy) => policy.workspaceAccountID === cardFundID);
	if (!matchingPolicy) {
	continue;
	}
	brokenCompanyCardConnections.push({
	policyID: matchingPolicy.id,
	policyName: matchingPolicy.name,
	cardID: String(card.cardID),
	});
	}
	}
	// Get personal cards with broken connections
	const brokenPersonalCardConnections: BrokenPersonalCardConnection[] = [];
	const personalCardsWithBrokenConnection = cardFeedErrors.personalCardsWithBrokenConnection;
	if (personalCardsWithBrokenConnection) {
	for (const card of Object.values(personalCardsWithBrokenConnection)) {
	brokenPersonalCardConnections.push({
	cardID: String(card.cardID),
	});
	}
	}

What changes do you think we should make in order to solve the problem?

Introduce a 90-day grace period at the derived layer only, so a single change suppresses both the task and the RBR while leaving the error itself intact.

1. Add a constant for the grace period (90 days), e.g. CONST.COMPANY_CARDS.BROKEN_CONNECTION_DISMISS_AFTER_DAYS = 90.

2. Add a helper in CardUtils — e.g. isBrokenConnectionPastDismissThreshold(card) — that returns true when the card's connection is broken and the connection has been unresolved for at least 90 days. Use the card's lastScrape timestamp as the anchor: it is the "Last updated" value already shown on the card detail pages and reflects the last successful data refresh (a separate lastImportAttempt field tracks attempts, so lastScrape does not move while the connection is failing). Its age therefore equals how long the connection has been broken. Compute the day difference with the existing date utility DateUtils.getDifferenceInDaysFromNow after parsing lastScrape, and compare against the constant. If lastScrape is missing, return false (fail safe — keep prompting).

App/src/types/onyx/Card.ts

Lines 99 to 109 in 826c670

	/** Last updated time */
	lastScrape?: string;
	/** Whether transactions from the card should be marked reimbursable by default */
	reimbursable?: boolean;
	/** Last update result */
	lastScrapeResult?: number;
	/** Last import attempt */
	lastImportAttempt?: string;

App/src/libs/DateUtils.ts

Lines 887 to 889 in 826c670

	function getDifferenceInDaysFromNow(date: Date) {
	return differenceInDays(new Date(), date);
	}

3. In cardFeedErrors.ts, inside both addErrorsForPersonalCard and addErrorsForCard, gate the local isFeedConnectionBroken so a card only contributes to the broken-connection signal while it is broken and within the 90-day window:

treat the connection as broken-for-surfacing only when isCardConnectionBroken(card) is true and isBrokenConnectionPastDismissThreshold(card) is false.

Because the broken-card task lists (cardsWithBrokenFeedConnection / personalCardsWithBrokenConnection) and the RBR aggregation (isFeedConnectionBroken / shouldShowRBR / shouldShowRbrForWorkspaceAccountID / shouldShowRbrForFeedNameWithDomainID) all derive from this one local, gating it in one place removes both the time-sensitive task and the RBR for connections unresolved past 90 days. Per-card OR-aggregation is preserved, so a feed with one card broken < 90 days still shows its RBR correctly.

4. Do not change isCardConnectionBroken itself, and do not touch card.errors. The card detail pages, the Fix-connection flows, the per-card error in the company cards table, and the wallet payment-method indicator all read the raw isCardConnectionBroken(card) / card.errors. Leaving those untouched is exactly what keeps "the error itself" — only the proactive home task and the navigational red dots are removed; the connection is still visible and fixable on its own card.

Net behavior: under 90 days broken → unchanged (task + RBR shown); 90+ days broken → home task and account/workspace/feed RBR removed, but the card still shows its broken state on its detail page and remains fixable.

Compare branch: https://github.com/Expensify/App/compare/main...wildan-m:App:wildan/91451-broken-card-90-day-grace?expand=1

What alternative solutions did you explore? (Optional)

• Modify isCardConnectionBroken to return false past 90 days. Rejected: that function is read directly by the card detail pages and the Fix-connection flows, so returning false would also hide the error itself and break the ability to fix the connection — violating the requirement.

App/src/pages/settings/Wallet/PersonalCardDetailsPage.tsx

Line 62 in 826c670

const isCardBroken = card ? isCardConnectionBroken(card) : false;

• Filter the 90-day check separately in TimeSensitiveSection and in each RBR consumer. Rejected: it duplicates the threshold logic across the home task, the account indicator, the workspace page, and the company cards table, which is easy to let drift out of sync. Gating the single derived source keeps the task and RBR consistent by construction.

• Wait for a backend-provided "connection broken since" timestamp. This could be swapped in later, but lastScrape's age is an accurate proxy today (no successful scrape happens while the connection is broken) and needs no API change, so the feature can ship frontend-only.

[melvin-bot]

Job added to Upwork: https://www.upwork.com/jobs/~022057928818599135150

[joekaufmanexpensify]

@jayeshmangwani my understanding is some backend changes will be required here, so going to source someone to assist with that.