shape-doctor.py members row: its green note says make siblings places every member with no working clone, a directory make siblings refuses included, so one row says both that it places IRRS and that it leaves IRRS exactly as it is #669

Goal: the members row says, member by member, what make siblings would do with the directory beside the holder, as the holder's scripts/siblings.py decides it. It never says make siblings places a member it will refuse.

Filed by lane openRepoShape-3 on Brett Heap's word of 2026-10-16T12:24:41Z ("give #460 to lane 3").

What happens:

• check_members puts every member whose sibling is not counted as its working clone into without_working_clone (shape-doctor.py:7574-7575). On the green path one note covers them all: "no working clone beside the holder for (make siblings places them when the holder is in a folder named after its repository, …)" (:7509-7518).

• working_clone turns a sibling down for several reasons, and make siblings refuses most of them (place and verify_sibling in templates/family-root/scripts/siblings.py):

  at ../<Project>	the doctor	make siblings	the note today
  nothing, or an empty folder	not counted	clones it	true
  a clone of the member from another remote (shape-doctor.py: the members row calls a beside-the-holder repository the working clone on its project id alone, where the family's own verifier also requires the origin #146)	other_origin set, stranger sentence	WRONG ORIGIN, left as it is (:484)	says both
  a non-empty folder with no .git	(None, None): is_this_member needs .git (:6929)	NOT A CLONE (:559)	false
  a repository that is not this member	(None, None)	WRONG ORIGIN (origin asked first) or WRONG PROJECT (:494)	false
  the holder itself (a member named after its family)	(None, None)	IS THE HOLDER (:548)	false
  a clone at the right origin with no readable project.yaml (read from the code, not run)	(None, None)	fetched as the member: project_id is None, so verify_sibling passes it	false
  a clone that fails at run time	cannot be known beforehand	CLONE FAILED (:585)	conditional; acceptable

• The contradiction is pinned by a test. R605 measured WRONG ORIGIN and NOT A CLONE on real holders. test_the_stranger_note_says_from_which_folder_make_siblings_calls_it (tests/test_shape_doctor.py:3703-3723) asserts that one row, for one WRONG ORIGIN member, holds both STRANGER_NOTE and PLACES_NOTE (:3720). This predates shape-doctor.py: the members row says make siblings places the working clones from a holder in a folder named after its repository, and that from anywhere else it clones nothing and prints the mv, as it does since #578 (#605) #624 (shape-doctor.py members row: once #578 lands, its note says make siblings places the clones, which it does only from a folder named after the family #605), which fixed the note's folder condition and left alone what the note is attached to.

• Neither the note nor the stranger sentence says the folder name must match EXACTLY. parent_folder_warning compares folder.name == family (siblings.py:969), so on macOS and Windows a folder inkrouter is refused for InkRouter. make siblings prints the rename itself.

• The verdict and exit code are right and must not move.

Where it is (d4a7fab):

• shape-doctor.py:
  • members_verdict, :7289-7521 (the stranger sentence at :7323-7329, the note at :7497-7518);
  • check_members, :7551-7578;
  • working_clone, :7128-7186;
  • member_state, :7189-7231;
  • is_this_member, :6913-6938.
• The rules it must agree with, in templates/family-root/scripts/siblings.py: place at :529-600, verify_sibling at :460-502, project_id at :449, and parent_folder_warning's exact compare at :969.

The change:

• Record what stands at ../<Project>. member_state takes the slot's state in place's order:

  1. the holder itself;
  2. a non-empty folder with no .git;
  3. absent or empty;
  4. a repository whose origin is not the member's;
  5. one at the right origin with another id;
  6. one at the right origin with no readable manifest;
  7. the working clone.

  The --json entry carries this state as a new key, so a caller need not parse prose.

• Say per member what make siblings does.

  • The "places them" note names only the absent and empty slots, and keeps its folder condition.
  • Every other member gets one sentence in make siblings's own word (WRONG ORIGIN, NOT A CLONE, WRONG PROJECT, IS THE HOLDER, or "fetched as the member") saying what it does. This holds on the green and the red path alike, as the stranger sentence does today.
  • CLONE FAILED is not predicted.

• Say "exactly". The folder condition reads "named exactly after its repository (case included)", in the note and in the stranger sentence.

• Document what the doctor restates. The doctor restates place's slot order but not the folder rule, as working_clone restates verify_sibling's questions (shape-doctor.py: the members row calls a beside-the-holder repository the working clone on its project id alone, where the family's own verifier also requires the origin #146). Say so in the docstring.

• Nothing else moves. without_working_clone keeps its meaning, and no verdict, exit code or NEXT changes.

Tests (tests/test_shape_doctor.py, the #146 and #605 sections, :3610-3725):

• Invert test_the_stranger_note_says_from_which_folder_make_siblings_calls_it: the stranger is not in the "places them" note.
• One row per slot state, each asserting the sentence, the --json key, ok and COMPLIANT.
• An agreement test: for each state, the holder's own scripts/siblings.py --dry-run, run from its family folder, gives the doctor's word for that member.
• PLACES_NOTE and STRANGER_NOTE gain "exactly".

From #460, #460 (comment) (L16, and the doctor's half of L21)

🤖 Generated with Claude Code