[bong-water-water-bong]

Fixes #1564. tools/copyright.py --check reds a developer clone on files that are not in the repository.

Cause, and why .gitignore is the honest fix

The gate lists tracked files plus untracked-but-not-ignored ones (#1414, so a new file cannot pass on a green it never earned) through:

listing("--others", "--exclude-standard")     # tools/copyright.py:171

--exclude-standard honours .gitignore — and nothing ignored .hrx123/, so three probe logs arrived as:

no comment style known (add one to LINE_STYLES/BLOCK_STYLES or EXEMPT): .hrx123/logs/probe-dumpcheck.jsonl

That is a red about the working tree, not about the repository: the shape that teaches people to skip the gate which caught base-modes-keep's duplicated licence block within two hours (#1551 → #1560).

Measured in a clean worktree of 15713661

state	--check
before, with .hrx123/logs/probe.jsonl present	rc=1, 1 line naming .hrx123
after .hrx123/ is ignored	rc=0, 0 lines naming it
untracked evil_probe.py with no header (the #1414 property)	rc=1, and the failure names the file
tests/copyright_check_untracked_test.py	0 failures, including "an ignored file is left out"

Why .gitignore and not an EXEMPT entry

.hrx123/ is entirely run output on this fleet — nothing under it is tracked on main (git ls-tree -r origin/main -- .hrx123 is empty) — and ignoring it also stops a log being committed by accident, which is how a dump/ tree would get in. EXEMPT would hide a run log from the gate even if someone committed it, which is the wrong side to err on; a path that must stay visible can be un-ignored deliberately.

One line plus a two-line comment in .gitignore; no code change, so the gate's behaviour on source is untouched.

[bong-water-water-bong]

@pi-prs ready for review at 81e16a9. Fixes #1564: one .gitignore entry (.hrx123/), because the gate already uses ls-files --others --exclude-standard. Measured before/after in a clean worktree: rc=1 with 1 named log -> rc=0; an untracked headerless source file still reds (the #1414 property); tests/copyright_check_untracked_test.py 0 failures.

[bong-water-water-bong]

Closing as a duplicate of #1568 — my mistake, and worth naming.

I claimed #1564 from a sweep of issues with zero comments and did not search open PRs for an existing fix first. #1568 (59bdc2be) had already put .hrx123/logs/ in .gitignore 13 minutes before my branch was pushed. That is precisely the duplicate the fleet was warned about tonight, and the check that would have caught it is one command I skipped:

gh pr list -R 1bit-MONSTER/engine --search "1564"

For the record, the only substantive difference in mine: I ignored .hrx123/ (the whole tree) where #1568 ignores .hrx123/logs/. I reproduced the same before/after (rc=1 with the named log → rc=0, while an untracked headerless .py still reds, and tests/copyright_check_untracked_test.py passes) — so #1568 does fix it, and the narrower path is defensible if dump/ is meant to stay visible.

Nothing of this branch is worth carrying: #1568 is the smaller, already-verdicted change. Commented there with the one-line thought rather than opening a competing diff.