Possible bug: PR still listed under "manual intervention needed", even though it's already on the backports branch

[DilumAluthge]

Consider the backports PR for 1.12.5:

• Backports for 1.12.5 JuliaLang/julia#60612

The "Contains multiple commits, manual intervention needed" section includes the following PR:

• Remove potential unitialized memory in the GC stack JuliaLang/julia#60651

However, if you look at the list of commits in the backport PR, you'll see the following: JuliaLang/julia@d6a0cd2:

From d6a0cd2009d6fa81c0739bb89b2a3ae38e2d6d14 Mon Sep 17 00:00:00 2001
From: Gabriel Baraldi <baraldigabriel@gmail.com>
Date: Fri, 16 Jan 2026 10:23:52 -0300
Subject: [PATCH] Remove potential unitialized memory in the GC stack (#60651)

(cherry picked from commit f9d461f4669d44dea86bed497b9498054d998988)

I did this cherry-pick manually earlier today. However, when I re-run the backporter script, the updated backports PR description (printed by the backporter script) still has the PR (JuliaLang/julia#60651) listed under "manual intervention needed". I've run the backporter script multiple times, but to no avail.

This isn't necessarily a huge problem, because the commit is on the backports branch, which is the important part. But there are some potential problems:

1. It's confusing for someone reading the backports PR description - they'll think that Remove potential unitialized memory in the GC stack JuliaLang/julia#60651 hasn't been backported, when in reality it has been.
2. This might also confuse @IanButterworth's label auditing tools? The correct behavior would be that after the backports PR (JuliaLang/julia#60612) is merged, the label auditing tool should correctly remove the backport 1.12 label from the PR (JuliaLang/julia#60651). But if the tool is confused (the same way the backports PR description generator is confused now), then maybe the backport 1.12 label won't get removed, which again might confuse the reader into thinking that the PR hasn't been backported to Julia 1.12.

Any idea why the backporter script doesn't think that JuliaLang/julia@d6a0cd2 is a backport of JuliaLang/julia#60651?

[IanButterworth]

#16