e2e-debug

You can debug frontend runtime bugs by orchestrating an inject-observe-cleanup workflow that inserts observable console.log probes into suspect source files, captures browser output, and removes all injections afterwards via a manifest-driven cleanup.

The best way to reproduce UI bugs in a browser session is to inject targeted console logs to reveal data shape mismatches, capture observed console output and network activity, apply a suggested fix, and remove all debug artifacts.

To ensure injected debug logs are removed after browser testing, a mandatory cleanup phase uses manifest-driven removal with a grep fallback to reliably strip all injected console probes from the source files.

Yes, you can resume an e2e debugging session using history capture designed for --continue workflows, allowing you to pick up persistent or ephemeral observer modes and iterative hypothesis loops where you left off.

Yes, the browser testing debugging approach supports automatic multi-round hypothesis loops alongside Teams persistent observer and subagent observation modes to systematically diagnose frontend runtime issues.

You should not use an inject-observe-cleanup workflow when your web application debugging requires no source file modifications, as this approach relies on directly injecting temporary console logs into suspect files to observe runtime behavior.