Playwright Run Diff

open source

See what changed between two test runs.

What it does

A green run and a red run rarely explain themselves. Run Diff takes two Playwright trace archives - a baseline and a candidate - and aligns them action by action, so the regression shows up as a list of findings instead of a guessing session.

Everything runs locally: parsing, matching and pixel diffing happen in a Web Worker in the browser app, or in Node through the CLI. Traces are never uploaded, and exported reports carry sanitized metadata only.

What is inside

  • Action alignment across two runs: added, removed, reordered and newly failing steps
  • Per-action timing regressions past absolute and relative thresholds
  • Network diffs: new or removed requests, status changes, latency deltas
  • Console diff: new errors, warnings and log-count changes
  • Pixel-level screencast diffs with configurable ratio and masks
  • Findings carry versioned rule ids, severity, confidence and investigation hints
  • Browser app and CLI with JSON, Markdown and self-contained HTML reports
  • Local-only processing - no upload, no account, CI-friendly exit codes

04From the product, not a mockup

What it looks like inside

Real screens from the live tool. Click any of them to expand.

More for Open Source

More for WordPress & WooCommerce

More for Wix

Have a workflow that deserves a tool?

Tell us what eats your week. If it is a real problem, we probably think about it already.

Talk to us