Quick question: Why does an error tracker need to block the main thread?

Quick question for fellow developers:

Why does capturing a frontend stack trace in 2026 still require 100KB+ of minified JavaScript and heavy fetch() retry queues that tank your Google Core Web Vitals (INP)?

When an app breaks, the monitoring tool shouldn't freeze the browser's main thread to serialize DOM breadcrumbs.

I got fed up with heavy APMs and built SnapTrace—a sub-5KB listener that fires asynchronously via navigator.sendBeacon (0ms delay) and collapses 50-error render loops into 1 alert.

You can test how it catches crashes and masks PII live in your browser (no signup):

👉 https://snaptrace-dashboard.vercel.app/test

Do you prioritize deep breadcrumb profiling, or do you prefer zero-bloat listeners that protect page speed?