A spent dispatch row keeps its issue for good, and 92 issues were held behind one #4440

What is wrong

A dispatch row that spends its retry ceiling is released once, after an hour,
and then stays rejected on purpose -- "a third identical failure is evidence
about the issue, not about the attempt". That rule is right, and its
consequence was never acted on: the issue stays claimed, so no bee can take
it again however much the brief improves.

Measured 2026-09-20: 101 open issues carried a queen-* branch, no pull
request and no running bee. 92 of them sat behind a branch that was empty or
had no merge base with master at all -- orphan debris from before the
container stopped cloning shallow. Every one had been held since 2026-09-17,
through two corrections to the brief they were written against (#4296, #4302).

The board showed it plainly: 59 cards in review, against 15 after the repair.

What to write

There is no external way to release a row. A NEW issue has no row, so copy the
work into one -- title, body and ## Boundary byte for byte, plus today's
toolbelt -- and close the old with a pointer.

It must refuse: a branch with real commits (that is work to publish), a branch
newer than the cutoff, an issue that has a pull request, an issue a bee is
running right now, and everything if the board does not answer.

Acceptance criteria

• 1. python3 tools/queen/refile.py --dry-run --limit 0 prints a line ending re-filed (today: no such file)
• 2. test -f .github/workflows/queen-refile.yml exits 0 (today: it does not exist)
• 3. python3 scripts/ci/check_pr_branch_filters.py prints a line beginning CLEAN: - the new workflow classified in the same commit
• 4. grep -c 'workflow files read 60' tools/census/quiet.txt prints 1 - and the census re-blessed in the same commit

Boundary

tools/queen/refile.py