build(deps-dev): bump vitest from 4.1.7 to 4.1.8 in the vitest group across 1 directory

[dependabot]

Bumps the vitest group with 1 update in the / directory: vitest.

Updates vitest from 4.1.7 to 4.1.8

Release notes

Sourced from vitest's releases.

v4.1.8

   🐞 Bug Fixes

• browser:
  • Disable client cdp API when allowWrite/allowExec: false [backport to v4]  -  by @​hi-ogawa and Codex in vitest-dev/vitest#10450 (e4067)
  • Remove orphaned Playwright route when same module is mocked via multiple ids [backport to v4]  -  by @​toxik and @​Zelys-DFKH in vitest-dev/vitest#10474 (675b4)

    View changes on GitHub

Commits

• e61f2dd chore: release v4.1.8
• e4067b3 fix(browser): disable client cdp API when allowWrite/allowExec: false [ba...
• See full diff in compare view

[baywet]

@dependabot recreate

[baywet]

This is impacted by this issue UmbrellaDocs/action-linkspector#62

And since we're using that action all over the place, it might start blocking other things. CC @lornajane @handrews @ralfhandl

[lornajane]

It looks like this pull request has commits from copilot and had also adjusted our usual merge rules (or I didn't understand something about the notifications I got, maybe?). We don't yet have an AI policy for our projects but I would not expect either commits or changes to merge rules to be performed by agents under the policy when we have it. I propose we therefore close this pull request and discuss further.

[handrews]

@lornajane agreed

@baywet can we roll back to a version of UmbrellaDocs/action-linkspector that does not have this problem?

[baywet]

@lornajane I initially thought the problem came from this update and had copilot try a couple of different fixes. But all the commits it created have been discarded. Only the one from dependabot remains.

@handrews yes we could I think. Unless the source of the problem is tied to a change in the underlying image, but I don't think so. I would prefer if it's fixed there so we don't have any additional work on our repos though. Let's give them a couple of days unless things are blocked and can't wait?

[baywet]

@handrews I've tried locking down the version to a hash from earlier a couple of times in #342 and none of them worked. I'm more convinced this is a "something changed on the agent image" issue. If the owners of the task don't reply, we should consider switching to something else.

[baywet]

@handrews I did iterate through the solutions linked to the issue, and #343 seems to do the trick :)