test: the private-key export screen has no e2e coverage, unlike the recovery-phrase screen #253

New Issue

clawbot · 2026-08-12T10:39:41+02:00

clawbot commented

2026-08-12 10:39:41 +02:00

The e2e suite's secret-screen cases are all show-phrase (from #161). export-privkey has none, even though #248 has just given it the same leave-hook and post-await liveness machinery.

Raised by the independent review of that PR and deliberately left out of its scope — its definition of done did not ask for e2e coverage, and popup init is already exercised where a ReferenceError in the touched views would surface.

Worth closing anyway: make check is prettier --check plus unit tests and cannot see a runtime error in a view handler, which is exactly how #150 and #151 shipped. The unit tests for this screen drive a stub DOM, so the real reveal-and-wipe path has never run in a browser.

Implementation requirements

Mirror the existing show-phrase e2e cases for export-privkey rather than inventing a new shape — the two screens now share their machinery, and the assertions should look the same.
Cover at minimum: the key is revealed after a correct password; leaving by the settings gear clears it from the DOM; leaving DURING the decrypt never puts it in the DOM.
Assert against the real rendered document, not a stub.

Definition of done

make test-e2e drives export-privkey through reveal and through both leave routes.
Removing the leave hook or the liveness guard makes the e2e run fail — demonstrated, with captured output.
TODO.md updated in the same commit.
make check passes.

The e2e suite's secret-screen cases are all `show-phrase` (from https://git.eeqj.de/sneak/AutistMask/issues/161). `export-privkey` has none, even though https://git.eeqj.de/sneak/AutistMask/pulls/248 has just given it the same leave-hook and post-await liveness machinery. Raised by the independent review of that PR and deliberately left out of its scope — its definition of done did not ask for e2e coverage, and popup init is already exercised where a `ReferenceError` in the touched views would surface. Worth closing anyway: `make check` is `prettier --check` plus unit tests and cannot see a runtime error in a view handler, which is exactly how https://git.eeqj.de/sneak/AutistMask/issues/150 and https://git.eeqj.de/sneak/AutistMask/issues/151 shipped. The unit tests for this screen drive a stub DOM, so the real reveal-and-wipe path has never run in a browser. ## Implementation requirements - Mirror the existing `show-phrase` e2e cases for `export-privkey` rather than inventing a new shape — the two screens now share their machinery, and the assertions should look the same. - Cover at minimum: the key is revealed after a correct password; leaving by the settings gear clears it from the DOM; leaving DURING the decrypt never puts it in the DOM. - Assert against the real rendered document, not a stub. ## Definition of done - [ ] `make test-e2e` drives `export-privkey` through reveal and through both leave routes. - [ ] Removing the leave hook or the liveness guard makes the e2e run fail — demonstrated, with captured output. - [ ] `TODO.md` updated in the same commit. - [ ] `make check` passes.

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: sneak/AutistMask#253