[miguel-heygen]

What

hyperframes browser clear reports a leftover browser folder that Windows still holds, instead of printing "Removed cached browser" while the folder is still there.

Why

#6771 made the shared leftover sweep skip folders Windows holds (EBUSY/EPERM), so a normal run is never blocked by a running Chrome's debug.log. browser clear goes through the same sweep, so it started swallowing the same error. A user who asks to clear the browser should be told when it did not clear.

Related work

Follows #6771, which this is a follow-up to. No other PR touches it.

How

The sweep takes a tolerateLocked flag, true for the commands that must keep running (download, render) and false for browser clear.

Test plan

• Unit tests added/updated
• Manual testing performed
• Documentation updated (if applicable)
• Comments follow CONTRIBUTING.md "Comments"

fails without the fix: with clearBrowser using the tolerant sweep again, does not fail a run when Windows holds a leftover's debug.log fails for both EBUSY and EPERM (2 failed, 71 passed).

What I measured

manager.test.ts: 73 passed on three runs on Linux; oxfmt and oxlint clean. The test now also covers EPERM and shows the sweep before a download and a forced re-download still resolve while browser clear throws.

What I did NOT execute

A real Windows machine with a running Chrome.

[github-actions]

Edit accuracy: accurate 271 (base branch 271), smooth 139 of those

The gate passes.
Smoothness is reported in the artifact, not gated. A case fails only if it fails 2 of 3 runs.

Quarantined, measured but not gated (0)

Unstable (1)

• seqplainkeys-none-pct-r0-root-z100: tracking 0.06, pressJump 0, drop 0, reload 0, render 0.03, renderKey -, undo false, teleport true / tracking 0.06, pressJump 0, drop 0, reload 0, render 0.03, renderKey -, undo true, teleport true / tracking 0.06, pressJump 0, drop 0, reload 0, render 0.03, renderKey -, undo true, teleport true

[heygengenesis]

Verdict: APPROVE · 0 blockers · 0 important · 0 nits · head 3626af06

I confirmed the fails without the fix: claim: with manager.ts set back to main, both does not fail a run when Windows holds a leftover's debug.log cases (EBUSY, EPERM) fail at clearBrowser(). With the fix, manager.test.ts passes 73/73, and oxlint and oxfmt are clean. Only clearBrowser passes false. The pre-download sweep, the forced re-download and the upgrade's set-aside removal keep the default true, so #6771's run-path tolerance is unchanged. Every Actions run at this head passed, including CI, Windows render verification and CodeQL.