← Back to Docs

Recipe: Error Tracking

Catch every crash, timeout, and anomaly before your users report it.

Overview

Meridian ships with a lightweight error ingestion pipeline that captures unhandled exceptions, promise rejections, and custom breadcrumbs. Everything lands in a single dashboard view with stack traces, device context, and session replay links.

Quick Start

import { Meridian } from '@getnimbus/meridian';

Meridian.init({
  dsn: 'https://ingest.getnimbus.net/proj/abc123',
  environment: 'production',
  sampleRate: 1.0,
});

Drop the snippet above into your entry file. Meridian automatically wraps window.onerror and unhandledrejection.

Manual Capture

try {
  await riskyOperation();
} catch (err) {
  Meridian.captureException(err, {
    tags: { feature: 'checkout' },
    user: { id: currentUser.id },
  });
}

Breadcrumbs

Attach context to every error by leaving a trail of breadcrumbs. Meridian records the last 100 events and attaches them to the next captured exception automatically.

Meridian.addBreadcrumb({
  category: 'auth',
  message: 'User logged in',
  level: 'info',
});

Dashboard Triage

  • Filter by release version, environment, or custom tag.
  • Mark issues as resolved, ignored, or assigned to a team member.
  • Set spike alerts: get notified when an error rate exceeds your threshold within a 5-minute window.

Pro tip: Combine error tracking with session replay to watch exactly what the user did before the crash.