[Fleet Monitor] combsmith/google-app-scripts — .github/workflows/playwright-tests.yml #651

Activity

1. don-petry

   added

   health-check

   dev-leadFor dev-lead agent pickupFor dev-lead agent pickup

   fleet-tracker

   on Oct 29, 2026
2. don-petry commented on Oct 29, 2026

   

   don-petry

   on Oct 29, 2026

   CollaboratorAuthor

   More actions

   Dev-Lead Implementation Plan

   Issue: #651 — [Fleet Monitor] combsmith/google-app-scripts — .github/workflows/playwright-tests.yml

   Root cause

   All 4 failing runs (40778439867, 40781501524, 40806734040, 40816061960) fail in a different deploy/tests/ui.spec.js test, always the same way: after signIn() + checking a script, #btn-deploy is disabled (click times out) or the deploy reports "select a script" (no .status-ok). Cause is a race in deploy/index.html: sign-in fires loadExistingDeployments() asynchronously, which calls renderScriptList() when the Drive lookup resolves. renderScriptList() rebuilds the checkbox HTML and wipes any selection the user made in the meantime, then setDeployEnabled() disables the button. This is a real UX bug (a fast user loses their selection), surfaced as test flakiness on CI.

   Scope

   Make renderScriptList() preserve the current checkbox selection across re-renders; add a deterministic regression test.

   Implementation Checklist

   • deploy/index.html: capture checked script ids before re-rendering and restore them in renderScriptList()

   Tests to Write

   • deploy/tests/ui.spec.js: hold the Drive files response, sign in, check a script, release the response, assert the checkbox stays checked and #btn-deploy stays enabled (fails on current code)

   Test Command

   npm run test:e2e (plus npm test, npm run check)