adopt-project.py execute: a hand split inside a directory the extension majority groups is recorded against the whole source's walk, so the doctor, which walks each leg, says MISPLACED on its first run after "adoption verified" #801

Goal: every adoption execute calls verified reads COMPLIANT on the doctor's first run. answered_paths: records each entry that the doctor's walk of THAT leg would dispute.

Filed by lane openRepoShape-1 from #457's line L1-83 (#457, comment) on Brett Heap's word of 2026-10-17T10:21:32Z, under lane 3's v1.1.1 rule (#473, issuecomment-6216697223). Measured on main bf1f099b with git-filter-repo a40bce548d2c and a stub gh.

Milestone: v1.1.1, not a regression. It is a wrong verdict reached by an ordinary hand split. This is L1-57's class, which #721 (#655) fixed for an override; here it is reached through a split instead.

Line

• L1-83. The source's engine/ holds a.py, b.py, c.py and x.md. That is 75% .py, exactly the policy's extension_majority: 0.75, so plan proposes engine/ → code. The person splits that entry by hand: engine/c.py → spec, and a.py, b.py and x.md → code, each with a resolution:.
  • check says plan ok.
  • execute --local-remote-dir exits 0 with "adoption verified".
  • shape-doctor.py --root <clone> --json exits 1 with verdict MISPLACED (1 path): code/engine/x.md: spec by rule extension-majority.

The cause. answered_path_record walks the whole source, where engine/ folds to code, so x.md needs no row. The doctor walks the code leg, where engine/ is 2 of 3 .py (67%). There it looks file by file, and x.md reads as spec.

Direction. In answered_path_record, also walk each leg's own file set, at the paths the doctor sees. Record each answered entry that covers a path this per-leg walk puts elsewhere. An unsplit plan still gets the same project.yaml, byte for byte. The doctor is unchanged.

Tests

• End to end, like test_an_adoption_answered_against_a_rule_is_compliant_on_first_run: the split above reads COMPLIANT, with the row ok.
• A unit test that answered_path_record records engine/x.md.
• Controls:
  • a split whose remainder still clears 75%;
  • an unsplit plan, whose project.yaml must be byte-identical.
• adopt-project.py and shape-doctor.py: a plan answer that overrides a confident rule passes check and execute, then the doctor's first run says MISPLACED, and an emptied .gitmodules is asked on every run #655's tests (tests/test_doctor_plan_overrides.py and tests/test_doctor_recorded_answers.py) still pass.

Lane: openreposhape-1 (openRepoShape-1)

🤖 Generated with Claude Code