The retained examples lean before contact; I had been looking mainly at the lifted view. Maintenance is comparing their seating with the ordinary housings under the inspection procedure.

That redirects the enquiry, but doesn't finish it. A raised incoming feature, a mismatched pocket and dirt beneath an insert can all leave a part leaning before pickup. What did the comparison actually find?

Emma's right. Don't rename arm drift as tray fault before the inspection comes back.

Quality found excess material on the underside of the affected incoming housings; the specified examples seat normally in the inspected pocket. The affected group is held. Saved pickup edits still need review before release of the job.

Does incoming inspection now have that underside feature in its check? Otherwise the same group can return in a different tray.

And retain the limits of the finding. Those identified housings explain the observed leaning examples, not necessarily every edit in the program's history. The reviewer should be able to leave older unexplained changes unresolved rather than invent a single story for all of them.