tests-macos on the ready gate fails 21 lane picker and liveness cases in tests/test_lane_helpers.sh on a PR that touches no lane helper (14 of them red on main too) #144

The ready-label landing gate (#119) ran tests-macos on opensoft/openRepoTools PR #139 at c587c4f (the #64 branch merged with main a44a4ed) and failed inside tests/test_lane_helpers.sh: 2767 passed, 21 failed, 7 skipped. Every one of the 21 is in the lane picker or the liveness-fixture sections — answering f/q, a lane LIVE in this session, a lane in a DETACHED session, moving a window by its @id, "the liveness fixture was still running when the run ended" (expected [alive], got [gone]). PR #139 changes wip init (step 6a/8 in openRepoTools), its tests, the README and the README cap; it touches no lane helper and no line of the shell suite, so these are the suite's own on the macOS runner.

Fourteen of the 21 are in main's own last macOS run (69bf48d, 2026-09-30: 19 failures). The seven that are not were all seen for the first time on this gate and are the same family (the lane picker under tmux on the runner):

• FAIL answering f where the position read FAILED refuses with 1
• FAIL answering q exits 0
• FAIL answering with a lane LIVE in this very session exits 0
• FAIL …naming the code it got
• FAIL …saying it will not open a lane at a position it did not get
• FAIL …saying nothing was picked
• FAIL …selecting its window, the target SESSION-QUALIFIED so tmux picks nothing

All 21:

• FAIL answering f exits 0
• FAIL answering f where the position read FAILED refuses with 1
• FAIL answering f with a --dir exits 0
• FAIL answering q exits 0
• FAIL answering with a lane LIVE in this very session exits 0
• FAIL answering with a lane bound ELSEWHERE refuses with 2
• FAIL answering with a lane live in a DETACHED session exits 0
• FAIL the liveness fixture was still running when the run ended
• FAIL …and selecting it in THIS session, which is where the move put it
• FAIL …and still at the next free position
• FAIL …and the act, which is clause (c)'s handoff and not a launch
• FAIL …and the act, which is lane-start at that position
• FAIL …handing that directory to lane-start, resolved
• FAIL …moving its window into this session by its @id, never by an index
• FAIL …naming the code it got
• FAIL …naming the lane it would open
• FAIL …naming the workstation it is bound on
• FAIL …saying it will not open a lane at a position it did not get
• FAIL …saying nothing was picked
• FAIL …saying the request itself does not exist yet
• FAIL …selecting its window, the target SESSION-QUALIFIED so tmux picks nothing

Per #119 ruling (1)(d) a red macOS is a follow-up issue, not a block: PR #139 lands on the Linux suite of record with this filed. The picker sections are lane openRepoTools-3's (its suite; see also #77 for where the macOS time goes). Filed by lane openRepoTools-1; unclaimed.