How to set up global error handlers in a web app without double-reporting

What should be true before you start?

Before adding handlers, document what your runtime and framework already report automatically. React, Next.js, and browser-level handlers can overlap in surprising ways once an SDK also captures uncaught failures.

• Check which global handlers the SDK already wires for errors and promise rejections.
• List the framework-level boundaries or custom handlers already present in your app.
• Decide where caught but important errors should call the manual helper instead of relying on global capture.

How should you implement this step by step?

The implementation pattern is simple: let one automatic path own uncaught failures, then build a small helper that dedupes manual capture for important caught errors. This keeps the signal complete without turning each exception into three incidents.

• Keep the SDK's automatic uncaught error and promise rejection capture as the default path for unexpected failures.
• Create one helper for manual reporting inside checkout, restore, export, or sync flows where caught errors still matter.
• Use a dedupe strategy so the same Error object does not get reported by multiple boundaries or handlers.
• Review grouped fingerprints after the first release to catch any remaining duplicate reporting patterns.

const seenErrors = new WeakSet()

function captureOnce(error) {
  if (error && typeof error === "object") {
    if (seenErrors.has(error)) return
    seenErrors.add(error)
  }

  Crossdeck.captureError(
    error instanceof Error ? error : new Error(String(error))
  )
}

Where do teams make mistakes?

Double-reporting usually starts as a harmless safety instinct, but it quickly turns into alert fatigue and misleading counts.

• Adding browser handlers, framework handlers, and manual capture without a dedupe policy.
• Re-wrapping every error into a new object and accidentally breaking grouping.
• Using global handlers as a substitute for targeted manual reporting in revenue-critical flows.

How does Crossdeck operationalize the workflow?

Crossdeck works best when the reporting contract is explicit: unexpected failures are automatic, important recovered failures are manual, and both land on the same customer timeline with clean grouping.

That structure keeps alerts useful and preserves trust in the dashboard, because the team can believe one incident really means one incident.

Frequently asked questions