Flaky or Fixable? is a browser triage game where you classify real failing tests pulled from open issues in Playwright, Cypress, and Selenium repos. For each card you see the test name, the failure output, and a few seconds of stack trace — then you decide: is this a flake, a real app bug, a user-error config issue, or a known framework issue you can park?
Use it as calibration training: drop it in front of a new QA hire to see how their gut compares to seasoned engineers, run a 10-minute round at a team meeting to surface shared blind spots, or use it as part of an interview loop for QA / SDET candidates.
Your final archetype score reflects triage intuition under noisy signals. Above 80% on the included issue set usually means your team has good failure-investigation habits and useful CI metadata. Below 50% is a signal that either the issues are genuinely ambiguous (welcome to QA) or your team's failure-classification heuristics need shoring up — usually fixable by adding more structured failure tagging to your CI pipeline.