Skip to content

test(console): a press on « Sign out » waits for My account at rest (FRU-248) - #293

Merged
sakuga-agent[bot] merged 1 commit into
mainfrom
fru-sf-entree/fix-e2e-sign-out
Oct 13, 2026
Merged

sakuga-agent[bot] merged 1 commit into
mainfrom
fru-sf-entree/fix-e2e-sign-out

Conversation

@sakuga-agent

@sakuga-agent sakuga-agent Bot commented Oct 13, 2026 •

Copy link
Copy Markdown
Contributor

main is red on 30c62f8 (job e2e): e2e/console-cookie-session.spec.ts:199, « ends the session on the worker when the person signs out », fails, retry included. This makes the test right. It is not a defect of the sign-out, and not a state shared between specs.

Cause, measured

The test opens My account and presses « Sign out » at once. My account (FRU-248, #269) draws its buttons, then reads the list « Where you are signed in », which arrives above « Sign out ». The button goes down 53 px. A press that starts before the list and ends after it is lost: the button has focus, no request leaves, and the page stays, with no message.

  • The page snapshot of the failed run shows exactly that: My account, the button « Sign out » [active], and no « You are still signed in ».
  • Reproduced each time: hold GET /console/sessions between the press and the release; the button moves from y=765 to y=818 and the address stays /account. With the list already there (control), the same press leads to /setup.
  • Four workers widen the window (the test takes 12 s in CI on four workers, 5.5 s on one). The merge of feat(console): the screen of an invitation: who invites, the role, « Join », then « You are in » (FRU-241) #270 does not touch this screen.

What changes

e2e/console-cookie-session.spec.ts only: its four presses on My account go through openAccount, which waits for the row of this browser in the list. No retry, no fixed wait, nothing run in series. console-account.spec.ts already waits for the list.

Proof

The two specs that open My account on a real worker, on four workers, four times each: 84 passed locally. Earlier, on main as it is, the four specs that start a worker passed 348/348 locally on four workers: the race needs the load of the CI to show, which is why the proof of the cause is the held request above and not a red run on my machine.

Not in this pull request

The shift is a defect of the screen too: a person can lose the same click. Reserving the room of the list in account.tsx is the next pull request, with the held request as its failing test.

🤖 Generated with Claude Code


Sakuga Review · Score de confiance : 5/5

Le changement ne touche que des tests e2e et peut être fusionné : il ajoute un helper qui attend que la liste des navigateurs soit affichée avant de cliquer sur « Sign out ».

Résumé

Le commit extrait dans un helper openAccount l'ouverture de My account suivie de l'attente de la région « Where you are signed in » contenant « Console · this browser », puis remplace l'appui sur le lien My account par cet helper dans quatre endroits du spec console-cookie-session. Le diff a été relu en entier ; le helper et ses quatre appels sont cohérents, et la locator getByRole/getByText utilisée existe dans le diff. Je n'ai pas pu vérifier la page My account elle-même ni le libellé exact dans l'application.

Non relu
  • e2e/console-cookie-session.spec.ts: les autres tests du fichier hors diff n'ont pas été revus
  • apps/console : le libellé « Console · this browser » et la région « Where you are signed in » n'ont pas été vérifiés dans le code de la page My account

Revues (1) · Dernier commit relu : test(console): a press on « Sign out » waits fo... · Niveau lite · anthropic/claude-haiku-5.5 · 0,001 $

…FRU-248)

main was red on 30c62f8: console-cookie-session.spec.ts pressed « Sign out » while My account still
read the list of the browsers. The list arrives above the button, which goes down 53 px; a press
that starts before and ends after is lost (the button has focus, nothing is sent, the page stays).
The page snapshot of the failed run shows that state, and holding GET /console/sessions between
the press and the release reproduces it each time.

The four presses of this spec now open My account through one helper, which waits for the list.
No retry, no fixed wait. The shift itself is a defect of the screen, fixed apart.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

@sakuga-review sakuga-review Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Score de confiance : 5/5 · ✅ Rien de bloquant · Le changement ne touche que des tests e2e et peut être fusionné : il ajoute un helper qui attend que la liste des navigateurs soit affichée avant de cliquer sur « Sign out ».

Niveau lite · anthropic/claude-haiku-5.5 · 0,001 $ · Le résumé, le diagramme et la liste des constats sont à la fin de la description de la PR.

@sakuga-review sakuga-review Bot added the review: 5/5 Score de confiance de Sakuga Review (5 : prête à merger) label Oct 13, 2026
@sakuga-agent
sakuga-agent Bot merged commit 3e194bf into main Oct 13, 2026
13 checks passed
@sakuga-agent
sakuga-agent Bot deleted the fru-sf-entree/fix-e2e-sign-out branch October 13, 2026 17:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

review: 5/5 Score de confiance de Sakuga Review (5 : prête à merger)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant