The bug-bash repro tests/bugbash/repros/editCancelKeepsDraft.e2e.ts failed once on the phone target, in the merge-group run https://github.com/coder/xum/actions/runs/37763109369 (job "Test / E2E (linux 1/4)", for #5875):
FAIL phone repros/editCancelKeepsDraft.e2e.ts > Cancel in edit mode leaves edit mode and keeps the unsent draft
ASSERTION_FAILED: expect.toHaveValue failed
locator: getByRole("textbox", name: "Message")
expected: value "unsent draft"
observed: value "" (match count 1)
❯ repros/editCancelKeepsDraft.e2e.ts:28:5
The same tree passed this shard on the PR run (https://github.com/coder/xum/actions/runs/37761591117). #5875 changes only tests/bugbash/sandbox/, and no repro imports those files. The failure is intermittent. It has two possible causes: a real intermittent loss of the draft when edit mode is cancelled at phone width, or a timing gap in the repro (the value check right after toBeHidden). I did not reproduce it locally. Among the 20 most recent failed pr.yml runs, only this run shows the failure.
Generated with xum • Model: anthropic:claude-opus-5-5 • Thinking: high • Cost: $115.24
The bug-bash repro
tests/bugbash/repros/editCancelKeepsDraft.e2e.tsfailed once on the phone target, in the merge-group run https://github.com/coder/xum/actions/runs/37763109369 (job "Test / E2E (linux 1/4)", for #5875):The same tree passed this shard on the PR run (https://github.com/coder/xum/actions/runs/37761591117). #5875 changes only
tests/bugbash/sandbox/, and no repro imports those files. The failure is intermittent. It has two possible causes: a real intermittent loss of the draft when edit mode is cancelled at phone width, or a timing gap in the repro (the value check right aftertoBeHidden). I did not reproduce it locally. Among the 20 most recent failedpr.ymlruns, only this run shows the failure.Generated with
xum• Model:anthropic:claude-opus-5-5• Thinking:high• Cost:$115.24