The sweep lists a leg declared with an absolute path as a paired tree of its own, keeps a leading ./ in a leg path, and lists dot-named feature directories, all of which lane-handoff and lane-reconcile skip or strip #276

Found by openRepoShape-2's independent review of #246 (T018) at 5dd0df07, for lane openRepoTools-1's landing gate (#246, comment 6266578777: verdict LAND, none of its findings blocking; this is its NIT 2). #246 landed as 55aae16. It still holds on main 00d52ae: lane-worktrees is unchanged between 55aae16 and 00d52ae, and the case below was run again on main's own lane-worktrees.

What happens

• The three readers of an assembly's paired root, <assembly>/worktrees/<feature>/<each declared leg path>, disagree on edge inputs.
• The sweep (lane-worktrees) joins every declared leg path under every feature directory, unfiltered, and lists every directory name. With Sweep._paired called directly, as the review did:
  • a leg declared with an absolute path becomes a FOREIGN paired row of the leg itself;
  • ./code keeps the ./ in the row's path (001-a/./code);
  • a feature directory named with a leading dot is listed.
• lane-handoff and lane-reconcile (in lanes-edit.sh) skip absolute paths, strip a leading ./, and, because they glob with *, skip dot names.
• Run again on main 00d52ae, with Sweep._paired of main's lane-worktrees called directly, as the review did: legs declared as code, ./spec and an absolute path to a leg checkout, features 001-a and .dotfeature, and a .git in each leg directory. The rows were the absolute-path leg (foreign, paired), .dotfeature/code, 001-a/./spec and 001-a/code. lanes-edit.sh project-legs hands the declared path over as written (it drops only an empty path, . and ./., and trims a trailing /: lanes-edit.sh:5521-5530 at 00d52ae), so the sweep does meet these forms from a real manifest.
• Weight: a NIT. The standard writes leg paths as relative ones (lanes-edit.sh:14287-14289), so a manifest that departs from it, or a dot-named directory under worktrees, is what meets it.

Where

• The sweep: lane-worktrees:2570 (decls) and :2585 (for full in [fdir] + [os.path.join(fdir, d) for d in decls]) in Sweep._paired, on main 00d52ae; the review's :2585 at 5dd0df07 is the same line.
• The other two: lane-handoff:2437 (the glob), :2445-2446 (case "$lg_decl" in ''|/*) continue ;; esac, ${lg_decl#./}); lanes-edit.sh:14281 (the glob) and :14291-14292 (awk $2 !~ /^\//, sub(/^\.\//, …)), all on main 00d52ae.

How it was found

• openRepoShape-2's independent review of Make the lane tools look inside a triad's legs and paired feature trees, so the sweep, lane-end's gate, lane-handoff, lane-reconcile and the daily report can see a lane's work there, and change nothing a checkout with no project.yaml does #246 at 5dd0df07 (tree ea4db0d): NIT 2, "the three readers of the paired root disagree on edge inputs", shown with Sweep._paired called directly.

The fix the review proposed

• Filter the sweep's list the same way: skip absolute leg paths, strip a leading ./, and skip dot names.

Why it is not in v1.1.0

• It is a NIT, and the review found none of its findings blocking the landing of Make the lane tools look inside a triad's legs and paired feature trees, so the sweep, lane-end's gate, lane-handoff, lane-reconcile and the daily report can see a lane's work there, and change nothing a checkout with no project.yaml does #246, which landed as 55aae16. The fix changes which rows the sweep lists.

Filed in milestone v1.2.0 on Brett Heap's word (2026-10-19T18:3xZ, "yes, file the v1.2.0 issues"). It is unclaimed.

Lane: openRepoTools-3

🤖 Generated with Claude Code