Tracking: adopt-project.py, converting an existing repository in place

Lane: openreposhape-3 (openRepoShape-3)

The parent issue for one area of openRepoShape: converting an existing repository into an assembly root with a spec leg and a code leg (adopt-project.py plan, check, execute). Brett Heap ruled on 2026-10-12 that the open issues are grouped under one parent per area, with a release milestone. Every open issue in this area is a sub-issue here.

v1.0.0. The first release is "core flows, all OSes": every procedure in AGENTS.md works end to end for ordinary repositories on Linux, macOS and Windows. The tag waits for every in-scope issue to close and for a full test run to pass on all three OSes on the release commit. In this area, the sub-issues in the v1.0.0 milestone are:

adopt-project.py execute exits 2 with nothing on stderr when git filter-repo fails, after both leg remotes exist: _build_the_legs swallows CommandFailed on the premise that run already reported it, and run only raises #171 (P2): A failure extracting a leg ends execute exit 2 with empty stderr, both leg repositories left behind; needs a git failure first; adopt-project.py: every failure once a leg repository exists flushes stdout, then prints REFUSED <step>., what failed and one notice of what the run left, so a failed git filter-repo no longer exits 2 with an empty stderr (#171, #188, #177) #445 fixes.
RULESET_HINT advises push -u origin ... from an extracted leg whose origin filter-repo removed, and says "listed above" of lines a combined log prints after it #177 (P1): Adopting into a ruleset org: the refused-push hint prints push -u origin, but filter-repo removed origin, so the advised command fails.
adopt-project.py: a failure AFTER both legs are pushed (the split-branch push, the mount, the split commit) prints only REFUSED <step> with no notice of what already exists and no stdout flush, so the combined log shows REFUSED before the bare … lines and the re-run meets leg-remote-exists unexplained #188 (P2): A split-push, mount or commit failure after both legs exist prints REFUSED but not what already exists, so the re-run's leg-remote-exists is unexplained.
adopt-project.py --pin <name> is accepted by plan, check and execute ("adoption verified"), but writes only the name into neutral_product_pins: and no contracts/<name>-pin.yaml, so the adopted root's own validate-pins.py (and its validate.yml) refuses with neutral-pin-missing from the first run #273 (P1): adopt-project.py plan --pin <name> (in docs/cli.md) ends adoption verified, then the adopted root's own validate-pins.py refuses neutral-pin-missing: red CI from day one.
adopt-project.py check passes a plan that keeps the source's own project.yaml in the root, and execute writes the shape's manifest beside it as shape/project.yaml and says adoption verified, so shape-doctor.py calls the adopted assembly NOT A SHAPE ROOT and its first make bootstrap fails #305 (P2): A source with its own top-level project.yaml answered root is plan ok, adoption verified, then NOT A SHAPE ROOT; spec/code/drop work.
adopt-project.py: plan, check and execute never look for the plan's split branch (adopt/three-repo-shape) in the repository the split is pushed to, so a re-run after an earlier run left it passes check, creates both legs, and dies at a non-fast-forward push of the split #320 (P1): AGENTS.md's recovery after a failed verification (re-run into fresh legs) meets the leftover adopt/three-repo-shape: check passes, both legs created, push rejected.
adopt-project.py plan: --spec-path and --code-path have no help=, so plan --help and docs/cli.md print two bare lines for the flags that choose where each leg is mounted, and say nothing of the two refusals plan makes on them before anything is created (adopt-leg-path-collides, adopt-bad-leg-path), which #288 is documenting for the scaffold's twins #340 (P2): plan --help and docs/cli.md print --spec-path/--code-path bare, saying nothing of the default or the two refusals; docs gap, no behaviour change.

Known limits. Every other sub-issue is a known limit: only exotic input triggers it (bytes that are not UTF-8, the C locale, control characters, hand-edited YAML, symlink tricks, names that differ only by case). The release notes list them, and they are worked when a release takes them in. A new edge case that review finds goes on this checklist, not into a new issue, unless it is in a release's scope or would hit a real user (Brett Heap, 2026-10-12).

Known limits measured by lane openRepoShape-3 on main 45e01f4 and kept here rather than filed:

→ adopt-project.py execute: under a non-UTF-8 locale, a moved or dropped path that is not ASCII kills git rm after both legs are pushed #527, done in adopt-project.py: execute hands git rm each path the split moves or drops as the bytes the tree holds, so a name outside ASCII under LC_ALL=C PYTHONUTF8=0 no longer stops the run with UnicodeEncodeError after both legs are pushed (#527) #614 (f6376d7) X1: under LC_ALL=C PYTHONUTF8=0, execute hands a moved or dropped path that is not ASCII to git rm as an argument subprocess cannot encode, so plan and check pass, both legs are pushed, and the run ends in a UnicodeEncodeError traceback. Ruled 2026-10-12: pass git the exact bytes the tree holds, under any locale.
-> adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900 A6: _split_message lists the root and dropped paths as they stand, so a name with a line feed, a trailing CR, ESC, a tab or a byte that is not UTF-8 (written as ?) reaches the split commit's message and the pull request body as another path or as raw control bytes. Direction: spell them as plan and check do.
→ adopt-project.py plan: ask about single-repository.yaml with its own drop-or-root rule, not the generic ambiguous-top-level-yaml question #528, done in contracts/path-classification.yaml: a source's own single-repository.yaml is asked drop or root, drop by default, under its own rule, so adopt-project.py plan no longer asks a record no leg reads the spec, code or root question (#528) #566 (1d809ea) A7: plan asks about a repository's single-repository.yaml with the general ambiguous-top-level-yaml question. Ruled 2026-10-12: a path-policy rule names the file and asks drop or root, defaulting to drop.
→ adopt-project.py check: refuse the branch names execute fails on after both legs exist (case, hideRefs, git's own rules) #520, done in adopt-project.py: check finds, and execute refuses before any leg, the branch names execute failed on, so adopt_branch: a..b, a source branch Adopt or one hidden by hideRefs no longer pass check and fail after both legs exist (#520, #420) #583 (de428f5) A4: adopt-project.py: check reports the split branch an earlier run left as execute's adopt-branch-exists, and a branch in git's way of it as plan-split-branch-taken, which execute now refuses before any leg exists, so a source branch adopt no longer has both legs made before git filter-repo fails (#320) #382's plan-split-branch-taken compares branch names case-sensitively, so on macOS and Windows a source branch that differs from the split branch only by case passes check, and execute makes both legs before git refuses. Ruled 2026-10-12: compare case-folded on every OS.
-> adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900 AS-1: _spelled writes a backtick as two characters the plan's reader keeps (\`) and a double quote unescaped, so a spelled name copied into the plan names another entry or makes the plan unreadable. Direction: \x60 and \".
-> adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900 AS-2: _shown prints a printable text that begins and ends with a backtick as it is, which is exactly its spelled form for another name, so plan's table can print two names as one. Direction: spell such a text too.
-> adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900 AS-3: plan's terminal report prints a path policy's question: and rule id as they are. Ruled 2026-10-12: print them through _shown, as check prints the question.

Known limits found by lane openRepoShape-3's acceptance passes (standard 36538fd, each re-checked on main 7a8b981) and kept here rather than filed:

→ adopt-project.py: remove the temp work roots it makes without --work-dir, keep a recorded --source clone, and point the tests' TMPDIR at their basetemp #526, done in adopt-project.py and tests/conftest.py: a run without --work-dir removes the empty work directory it made, and a recorded clone that is gone is cloned again, so no empty openreposhape-adopt-* is left and no source-unclonable follows (#526) #594 (c972032) plan, check and execute each make a fresh openreposhape-adopt-XXXXXXXX directory in the system temp folder when --work-dir is not given (_work_root, adopt-project.py:6028-6032, called at :2088, :3543 and :5100) and nothing ever removes it: plan and check of a local source leave it empty, execute leaves both legs' and the split's working trees in it, and 2,992 of the first 3,000 such directories in one container's /tmp (15,206 in all) are empty. Direction: remove a work root that holds nothing when the command ends, and say where execute's stays.
→ adopt-project.py: what execute and AGENTS.md say around a run (the plan's follow-up, a failed verification's open PR, an answer with no resolution:, the ruleset hint) #523, done in adopt-project.py: what plan, execute and AGENTS.md say around a run is what the run does: the plan's last follow-up is a review of the pull request execute opens itself, a failed verification names the pull request and the topic it left in place, AGENTS.md says check passes an answer with no resolution:, NEXT under --local-remote-dir names the split's branch, and the ruleset hint no longer calls a seed branch and a pull request the scaffold's other exit (#523) #582 (526cd2c) AT-adopt N1: the follow-up "open the split as a pull request against ...; the legs -spec and -code are seeded before it" (adopt-project.py:1368-1372) is printed by plan and listed under "FOLLOW-UPS, which this commit does NOT do" in the split commit and the pull request body (:5442-5444), though in GitHub mode execute opens that pull request itself, and it calls two legs "seeded" that were extracted (the docs use that word only for a leg made from a template).
→ rides with adopt-project.py: what execute and AGENTS.md say around a run (the plan's follow-up, a failed verification's open PR, an answer with no resolution:, the ruleset hint) #523, done in adopt-project.py: what plan, execute and AGENTS.md say around a run is what the run does: the plan's last follow-up is a review of the pull request execute opens itself, a failed verification names the pull request and the topic it left in place, AGENTS.md says check passes an answer with no resolution:, NEXT under --local-remote-dir names the split's branch, and the ruleset hint no longer calls a seed branch and a pull request the scaffold's other exit (#523) #582 (526cd2c) AT-adopt N2: under --local-remote-dir no pull request is opened, yet execute ends with NEXT: review the pull request on <Project>, then git clone ... make bootstrap (adopt-project.py:5190, _next_steps at :5577; adopt-project.py: the tracking branch defaults to the source's default branch, topics are set before the pull request, and NEXT names scripts/bootstrap.py when the shape's Makefile went aside (#479, #480, #481) #493 keeps it, its condition is refused is None). Test path only.
→ adopt-project.py: what execute and AGENTS.md say around a run (the plan's follow-up, a failed verification's open PR, an answer with no resolution:, the ruleset hint) #523, done in adopt-project.py: what plan, execute and AGENTS.md say around a run is what the run does: the plan's last follow-up is a review of the pull request execute opens itself, a failed verification names the pull request and the topic it left in place, AGENTS.md says check passes an answer with no resolution:, NEXT under --local-remote-dir names the split's branch, and the ruleset hint no longer calls a seed branch and a pull request the scaffold's other exit (#523) #582 (526cd2c) AT-adopt N3: AGENTS.md:134-136 says an answer with no resolution: beside it "will be read as a finding", but check prints plan ok, exit 0, for one (the record's resolution: is optional, adopt-project.py:4992-4995). Direction: say who reads it, or let check ask for one.
Closed 2026-10-14: a duplicate of auA-12, now in adopt-project.py: what execute and AGENTS.md say around a run (the plan's follow-up, a failed verification's open PR, an answer with no resolution:, the ruleset hint) #523. AT-adopt N5: when the verification fails in GitHub mode (exit 1) the pull request and the topics are already done, yet the report and AGENTS.md:147-150 never name the open pull request; the way out they print (delete the split branch, re-run into fresh legs) does work.
→ adopt-project.py plan: with no --elected-by and no user.name, the refusal reads --elected-by is '', naming a flag that was never passed #529, done in adopt-project.py: plan names an elector it took from git config user.name by where it came from, so with no --elected-by and no git config user.name the refusal reads --elected-by (not given, so git config user.name) is '' and no longer --elected-by is '', a flag nobody passed (#529) #567 (9f80939) AT-adopt N6: plan with no --elected-by and no git config user.name is refused adopt-manifest-invalid: --elected-by is '', naming a flag the person never passed rather than the setting the empty default came from (probe on main 7a8b981; the label is adopt-project.py:1856, while the unwritable-value path at :1808 does say "not given, so git config user.name").
→ adopt-project.py: record the plan's answers in project.yaml, so the doctor's placement row and the submodules: rows follow them #525, done in adopt-project.py, shape-doctor.py, validate-manifest.py: execute records the plan's answers under answered_paths: and a moved submodule names its row, so an answered examples/ is not asked again and a deliberate move names no plan (#525) #617 (59e5a07) AT-adopt N7: the doctor's placement row keeps listing a path the adoption plan already answered (spec/examples/, rule ambiguous-examples) as review_required (shape-doctor.py:3544, :3831), because execute records only the submodule answers in project.yaml (submodule_record, adopt-project.py:4915-4950), so AGENTS.md's doctor step 4 has an assistant ask the human again. Direction: record a plan's answers for ambiguous paths as An adopted assembly records nothing about which leg a human chose for a source submodule, so shape-doctor.py cannot verify a kept gitlink's placement and can only note it; adopt should write the plan's submodule answers into the assembly #189 did for submodules.

Edge cases found in review of lane openRepoShape-3's pull requests, kept here rather than filed:

→ adopt-project.py check: refuse the branch names execute fails on after both legs exist (case, hideRefs, git's own rules) #520, done in adopt-project.py: check finds, and execute refuses before any leg, the branch names execute failed on, so adopt_branch: a..b, a source branch Adopt or one hidden by hideRefs no longer pass check and fail after both legs exist (#520, #420) #583 (de428f5) adopt-project.py reads the branches of a local source or assembly with git ls-remote (_branches_at, adopt-project.py:3830, called at :3974, :4028 and :4034 by adopt-project.py: check reports the split branch an earlier run left as execute's adopt-branch-exists, and a branch in git's way of it as plan-split-branch-taken, which execute now refuses before any leg exists, so a source branch adopt no longer has both legs made before git filter-repo fails (#320) #382's _places_check_asks and _refuse_a_branch_in_the_split_branchs_way), so a branch the repository hides with uploadpack.hideRefs or transfer.hideRefs is not seen: check says plan ok, and execute --local-remote-dir pushes both legs before the split push fails on git's ref lock. Direction: read a local path's branches with git for-each-ref. From the review of adopt-project.py: check reports the split branch an earlier run left as execute's adopt-branch-exists, and a branch in git's way of it as plan-split-branch-taken, which execute now refuses before any leg exists, so a source branch adopt no longer has both legs made before git filter-repo fails (#320) #382 (codex at 5a1ffd5, confirmed by an independent verification; here only the git side was re-checked: git ls-remote prints nothing for a transfer.hideRefs branch that git for-each-ref lists, and execute was not run, there being no git-filter-repo in this container).

(lane openRepoShape-1, 2026-10-18T08:08:32Z) Two limits that lane 1's v1.1.1 PRs named for this checklist: L1-104 and L1-105.

Each was found in its PR's review and left out of it, as its squash says. The v1.1.1 release notes (#883) list both under known limits as on this checklist, and lane 3's review of #883 found them missing here (#883, issuecomment-6247451611).

From #829 (#801, dd72294d), the extension-majority fold:

[L1-104] A tracked name that the doctor sets aside before it walks a leg, one holding a control character or a byte that is not UTF-8, still counts in plan's and execute's extension-majority fold. So engine/ holding a.py, b.py, x.md and t<TAB>b.py folds to code, 3 of 4, and the doctor, which sees 2 of 3, reads code/engine/x.md as MISPLACED, with the plan exactly as plan wrote it. It is true on main before adopt-project.py: answered_paths: also records each entry a leg's own walk puts elsewhere, so a directory split by hand out of an extension-majority fold no longer reads MISPLACED on the doctor's first run after adoption verified (#801) #829 as well; its cause is the doctor's set-aside, not a split.

From #831 (#796, 33419925), --source:

[L1-105] A --source directory this user cannot enter (chmod 000), or one under a directory they cannot enter, still ends adopt-project.py in a PermissionError traceback, exit 1, as on main before adopt-project.py: a --source, or a plan's source.local_path, that names no repository's top level is refused by name before git reads anything else in it, so a file, a directory inside a repository, an unknown ~user, an empty value, owner/.git and a lost blob no longer end in a traceback, plan a subtree or reach gh (#796) #831: from git rev-parse in _top_level for the first, and from Path.exists in _source_spec for the second.

Lane: openreposhape-1 (openRepoShape-1)

docs/release-notes/v1.1.1.md: the v1.1.1 notes take RV883's NIT list …

(lane openRepoShape-3, 2026-10-18T14:46:39Z) Triage of #457's unchecked items at the v1.1.1 release commit 5017830. This is lane 3's TRIAGE-457, read only, before this filing.

Counts. 144 lines were unchecked, 9 in the body and 135 in comments. The rulings:

FIXED: 45.
FIXED in part: 1 (L1-82).
STALE: 1 (L1-103).
OPEN, filed: 53, in twelve issues in milestone v1.2.0, adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900 to adopt-project.py and family.py: a --source this user cannot enter, a UTF-8 plan checked under the C locale, and a member branch that is not UTF-8 still end in a traceback or a false refusal #918.
OPEN, not filed: 40 known limits that the v1.1.0 or v1.1.1 notes already rule.
UNSETTLED: 4.

No open issue or pull request covered any of them. Every -> pointer on this checklist names a closed issue.

How. Read: this issue's body and every comment, the release notes' Known limits sections, and the first-parent squash messages on main. Probes ran on a detached clone at 5017830, with:

a stub gh first on PATH (exit 97), GH_TOKEN=bogus and an empty GH_CONFIG_DIR;
GIT_CONFIG_GLOBAL=/dev/null, and local remotes only.

No pytest was run. Each OPEN row names its issue, whose body holds the probe output.

Filed:

adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900: A6, AS-1, AS-2, AS-3, L2-76, L1-36.
adopt-project.py check and execute: one bad plan value is named in different orders, or by two or three findings, and two pin sentences say more than execute does #901: L2-6, L2-7, L1-K7, L2-12, L2-42.
adopt-project.py plan and check: an empty flag or a bare plan value is taken without a word, and the id refusals name what the person did not type #902: L1-K3, L2-86, L2-87, L2-88, L2-89, L2-90, L1-49.
adopt-project.py plan: refusals that need only the flags come after the work directory, the source's git reads and gh repo clone #903: L2-55, L4-37, L1-67.
adopt-project.py check: a root-kept symlink into the clone's .git, or through a kept link to a declared pin's file, passes, or is refused for a reason that is not true #904: L1-J1, L1-J2, L1-K2, L1-K9, with L1-K8 as its rider.
adopt-project.py: five printed sentences say what the run does not do: the Makefile follow-up beside a root Makefile, the contracts/-only read-across path, execute would stop there, a path names no attribute, and WHAT WILL HAPPEN under --local-remote-dir #906: L2-72, L2-73, L1-91, L1-93, L1-75.
adopt-project.py check: a symlink target with a NUL after its first byte, or longer than the platform takes, passes check and fails after both legs exist #908: L1-76, L1-78, L1-80.
adopt-project.py check's throwaway repositories: a SHA-256 source, a tree path git refuses, a TMPDIR on another disk and this machine's hooks change the answers #910: L1-90, L1-92, L1-94, L1-32.
adopt-project.py and shape_materialize.py: a tag whose target the clone lacks is called a tag that names no commit, a recorded tag id is asked only of a clone still in place, and a pinned product only PyYAML misreads is written bare #912: L1-81, L1-82 (second case), L2-14.
adopt-project.py and shape-doctor.py: a .gitmodules of whitespace is asked on every run, a dropped file whose blob the root also holds counts as root, _answered_legs is quadratic, and two validator code families have no placeholder row #914: L1-84, L1-85, L1-87, L1-K4.
tests: five adopt tests are weaker than they read: text on stdin, a stub gh written with write_text, GIT_ALLOW_PROTOCOL not cleared, two guard pairs that mask each other, and one name shape in the naming test #916: L4-54, L1-60, L1-62, L1-K6, L2-91.
adopt-project.py and family.py: a --source this user cannot enter, a UTF-8 plan checked under the C locale, and a member branch that is not UTF-8 still end in a traceback or a false refusal #918: L1-105, L1-86, L1-70.

This issue's body was edited once:

A7 and AT-adopt N1, N2, N3 and N6 are ticked, each with the change that fixed it.
A6, AS-1, AS-2 and AS-3 now point at adopt-project.py: a kept or dropped name reaches the split commit and the pull request body unspelled, and _spelled, _shown and plan's report print text that reads as another name, or raw #900.

Lines in comments stay as their authors wrote them. This comment records their rulings, and each author, or lane 3 on a later word, can tick them.

Body

Line	Ruling	Evidence
A6: split message and PR body list names raw	OPEN -> #900	probe: a kept `caf\xe9.md` lands as `caf?.md`, `cr\r` as `cr`; `adopt-project.py:13498-13509`
A7 (-> #528)	FIXED	`1d809ea` (#566, #528); `contracts/path-classification.yaml:453`
AS-1: `_spelled` leaves `"` raw	OPEN -> #900	probe: `_spelled('a"b')` is `a"b`; `adopt-project.py:3114-3124`
AS-2: `_shown` of a backticked text	OPEN -> #900	probe: `_shown` of `x` equals `_spelled('x')`; `:3111`
AS-3: question and rule id printed raw	OPEN -> #900	`:5168`, `:5173`, `:5495` (read)
AT-adopt N1	FIXED	`526cd2c` (#582, #523); `:3323`
AT-adopt N2	FIXED	`526cd2c` (#582, #523); `:13208`
AT-adopt N3	FIXED	`526cd2c` (#582, #523); `AGENTS.md:139-140`
AT-adopt N6	FIXED	`9f80939` (#567, #529); `:4033`

Comments, 2026-10-12

Line	Ruling	Evidence
L2-6: bad id or org beside a bad leg path	OPEN -> #901	probe: `check` refuses `plan-bad-id` (or `unsafe-value`), `execute` refuses `plan-bad-leg-path`, 0 legs
L2-7: `_manifest_values` docstring	OPEN -> #901	`:4911-4912`
`CONTRACTS_DIR` hard-coded (lane 5, -> #522)	FIXED	`fbbbc8d` (#572, #522); `scripts/shape_materialize.py:1045`
Windows x4 (lane 4, = L4-05 to L4-08)	known limit, not filed	skips remain (`POSIX_ONLY`, `tests/test_adopt_entry_paths.py:51`); v1.1.0 notes
auA-12 (its copy L4-09 is ticked)	FIXED	`526cd2c` (#582, #523); `:200`
auA-13 (its copy L4-10 is ticked)	FIXED	`fbbbc8d` (#572, #522); `:3224`
root `.gitattributes` is a symlink (= L4-15)	known limit, not filed	`scripts/shape_materialize.py:909-913`; v1.0.0 notes
older adoption has no block (= L4-17)	known limit, not filed	`extend_eol_block`, `scripts/shape_materialize.py:796-798`; v1.1.0 notes
L1-J1: link into `.git` where the clone holds nothing	OPEN -> #904	probe: `CLAUDE.md -> .git/refs/remotes/origin/main` passes `check`; `:8720`
L1-J2: `.git/refs/../HEAD` refused for a false reason	OPEN -> #904	probe: "passes through `.git/refs`, which is not a directory in the assembly root"

Comments, 2026-10-13 and 2026-10-14

Line	Ruling	Evidence
L2-12: `pins:` spelled with `repr`	OPEN -> #901	`:554`
L2-14: names only PyYAML misreads written bare	OPEN -> #912	`_neutral_pin_entry`, `:13393-13405`
#499 line: `adopt-not-the-source` after `gh repo clone` (= L4-37)	OPEN -> #903	probe: `gh repo clone` runs, then `source-unclonable`
#499 line: `check` gives no rename-first remedy (= L4-38)	FIXED	`d7befb9` (#825, #716); `:14452`
#499 line: recorded repository not consulted (= L4-40)	FIXED	`d7befb9` (#825, #716); `:14520`
#499 line: `casefold` (= L4-43)	known limit, not filed	`:14398`; v1.1.0 notes
#499 line: `_origin_path` vs `_remote_repository` (its copy L4-44 is ticked)	FIXED	#591 (`55faad2`, #521); `_origin_path` reads `_origin_url`, `:14430`
#499 line: `None/` org (= L4-45)	known limit, not filed	`:8446-8447`; v1.1.0 notes
#499 line: `_origin_path` on every plan (its copy L4-46 is ticked)	FIXED	#591 (`55faad2`, #521); called only at `:14440`, `:14553`
#499 line: comment wording on `org:` (= L4-48)	FIXED	`d7befb9` (#825, #716); `:14600`
L1-K2: pin file through a kept link	OPEN -> #904	probe: `contracts -> governance` passes `check`
L1-K3: bare value made text	OPEN -> #902	probe: `id: 007` -> `adoption verified`, `id: "7"`
L1-K4: no placeholder row for the submodule codes	OPEN -> #914	`scripts/shape_materialize.py:4935-4940`
L1-K6: guards that mask each other	OPEN -> #916	`:4120`, `:4264`, `:4161`
L1-K7: tab in a pin product	OPEN -> #901	probe: now three findings
L1-K8: `.git/reftable`	UNSETTLED, rider in #904	git 2.43 here; `CLONE_GIT`, `:8715`
L1-K9: false "after both legs" sentence	OPEN -> #904	`:6242-6258`; `tests/test_adopt_root_shape_paths.py:339`
`.github -> ci` block spelling (= L4-51)	FIXED	`d6475f9` (#844, #807); `scripts/shape_materialize.py:887`
hash-object fed text (= L4-54)	OPEN -> #916	`tests/test_adopt_ignored_shape_files.py:294-295`

Comments, 2026-10-15

Line	Ruling	Evidence
L2-42: pin finding's "and nothing else"	OPEN -> #901	`:584-586`
L2-43: `local-clone (<path>)` remedy	FIXED	`396c0ff` (#570, #540); `:503`; probe: `check` passes such a plan
L1-01	FIXED	`2d98820` (#725, #651); `scripts/shape_materialize.py:3489`
L1-02	known limit, not filed	`git_out`, `scripts/repo_shape.py:1206-1209`; v1.1.0 notes
L1-04	FIXED	`bee84d5` (#842, #800); `scripts/shape_materialize.py:3214`
L1-08, L1-10, L1-13, L1-14, L1-17, L1-18	known limits, not filed	v1.1.0 notes; L1-10's variable is held back on purpose (`:9544-9546`); L1-17 narrowed by #825
L1-29, L1-30, L1-31	known limits, not filed	v1.1.0 notes
L1-32: only `OSError` around the probe	OPEN -> #910	`:9374`
L1-36: C-locale table shift	OPEN -> #900	probe: the row shifts 3 columns
L1-38: macOS NFD, Latin-1	UNSETTLED	needs macOS or a Latin-1 locale
L1-39	FIXED	`b618ec1` (#699, #658); `:3835`
L1-40, L1-41, L1-42, L1-44, L1-45, L1-46, L1-47, L1-48	known limits, not filed	v1.1.0 notes; probe of `url_repository` at `5017830` confirms L1-42, L1-44, L1-46
L1-49: AGENTS.md half	OPEN -> #902	`AGENTS.md:119-130`
L1-50 to L1-54	known limits, not filed	v1.1.0 notes
L1-60: stub gh via `write_text`	OPEN -> #916	`tests/test_adopt_pins.py:112-122`
L1-62: `GIT_ALLOW_PROTOCOL` not cleared	OPEN -> #916	`tests/test_adopt_submodule_url_answers.py:301`
L2-55: flag-only refusals late	OPEN -> #903	`_work_root` `:5387` before `_checked_plan_inputs` `:5403`

Comments, 2026-10-16 and 2026-10-18

Line	Ruling	Evidence
L2-72: Makefile follow-up beside a root Makefile	OPEN -> #906	`:3346-3351`
L2-76: raw `git-failed`	OPEN -> #900	`:1593-1595`
`check` names one bad `tracking_branch` twice (= L4-127)	FIXED	`64ec47c` (#817, #708); `:5002`
L2-73: only `contracts/` gets a path	OPEN -> #906	`:3283`
L4-05 to L4-08, L4-15, L4-17, L4-43, L4-45	known limits, not filed	as their originals
L4-37 (its `-> #652` is stale)	OPEN -> #903	as its original
L4-38, L4-40, L4-48	FIXED	`d7befb9` (#825, #716)
L4-47	FIXED	`190133d` (#630); `README.md:776-777`
L4-51	FIXED	`d6475f9` (#844, #807)
L4-54	OPEN -> #916	as its original
L4-127	FIXED	`64ec47c` (#817, #708)
L1-64, L1-77, L1-88	FIXED	`3341992` (#831, #796); `:1839`, `:2237`
L1-65	FIXED	`5c02a04` (#843, #803); `:1894`
L1-66b: `mkdtemp` with no temp directory	UNSETTLED	probe: a missing `TMPDIR` falls back to `/tmp`, exit 0; the bare `mkdtemp` is `:14700`
L1-67: three nits	OPEN -> #903	probe: nit 1 holds; nit 3 by `:5387`; nit 2 not measured
L1-68, L1-69, L1-71	FIXED	`51414d2` (#830, #797); `:2670`, `:2539`
L1-70: `_member_head` strict decode	OPEN -> #918	`scripts/family.py:7188-7206`
L1-72, L1-73	FIXED	`fe99b8b` (#861, #798); `:11535`, `:313`
L1-74: `check` vs `execute` under `--local-remote-dir`	UNSETTLED	`adopt-name-too-long` now found by `check` (`a249910`, #871); the rest not re-measured
L1-75: "(private)" for bare repositories	OPEN -> #906	`:8579-8580`
L1-76, L1-78: late NUL, long target	OPEN -> #908	`:7710-7711`
L1-79	FIXED	`3fb30a8` (#833, #799); `:5453`
L1-80	OPEN -> #908	`_link`, `:8964-8977`
L1-81: tag whose target the clone lacks	OPEN -> #912	`scripts/shape_materialize.py:3498-3508`
L1-82	first case FIXED (`bee84d5`, #842, #800); second OPEN -> #912	`:829-846`
L1-83	FIXED	`dd72294` (#829, #801); `:12874`
L1-84, L1-85, L1-87	OPEN -> #914	`shape-doctor.py:246`; `:6354`; `:13830-13839`
L1-86: C locale over a UTF-8 plan	OPEN -> #918	probe: `plan-source-gone` for a path that exists, no `gh` call
L2-86 to L2-90	OPEN -> #902	probes in #902
L2-91: naming test asks one name shape	OPEN -> #916	`tests/test_naming_policy.py:252-254`
L1-89	FIXED	`3fb30a8` (#833, #799); `:5193`
L1-90, L1-92, L1-94	OPEN -> #910	`:10160`, `:10177`, `:9551`
L1-91, L1-93	OPEN -> #906	`:10444-10448`, `:9675`
L1-95, L1-96, L1-100, L1-101	FIXED	`3341992` (#831, #796); `:1770`, `:1712`, `:1680`
L1-97, L1-98, L1-99, L1-102	FIXED	`5c02a04` (#843, #803); `:1960`; L1-102 not run on Windows, and its residual is in v1.1.1's notes
L1-103	STALE	ruled by design in `60f146a` (#749): `execute` keeps its clone, and a re-run is refused `adopt-work-dir-occupied`
L1-104	known limit, not filed	v1.1.1 notes
L1-105	OPEN -> #918	probe: chmod 000 gives a `PermissionError` traceback; also in v1.1.1's notes

File:line references without a file name are in adopt-project.py, at 5017830.

Not filed.

Known limits that the release notes already rule. Each is filed only if a release takes it in:
- how an origin is read: L4-43, L4-45 (and their originals), L1-40, L1-41, L1-42, L1-44, L1-45, L1-46, L1-47 and L1-48;
- L1-02, L1-08, L1-10, L1-13, L1-14, L1-17, L1-18, L1-29, L1-30, L1-31, L1-50, L1-51, L1-52, L1-53 and L1-54;
- the four Windows lines (L4-05 to L4-08, and their originals), L4-15 and L4-17 (and their originals), and L1-104.
UNSETTLED:
- L1-K8. It needs git 2.45 or later; it rides on adopt-project.py check: a root-kept symlink into the clone's .git, or through a kept link to a declared pin's file, passes, or is refused for a reason that is not true #904.
- L1-38. It needs macOS or a Latin-1 locale.
- L1-66b. It needs a host with no usable temporary directory, or a test that makes mkdtemp raise.
- L1-74. It needs the rest of its cases run under --local-remote-dir.

Housekeeping.

L4-15's -> #654 and L4-37's -> #652 are stale: both issues closed without the line. L4-15 is a known limit, and L4-37 is now in adopt-project.py plan: refusals that need only the flags come after the work directory, the source's git reads and gh repo clone #903.
Lane 4's comment of 2026-10-16T14:10:49Z copied its lines word for word as L4-05 to L4-17, L4-35 to L4-54 and L4-127, so 17 originals in its earlier comments are still unticked beside their copies:
- the four Windows lines;
- auA-12 and auA-13;
- the root .gitattributes symlink and the older-adoption lines;
- the eight unticked adopt-project.py plan never compares --project (or --org) with the repository it adopts in place, so a source named billing-service is refused with a remediation that steers to --project BillingService, which plans and checks plan ok against acme/billing-service, and execute then acts on acme/BillingService: a raw git ls-remote failure when no such repository exists, and, when a different repository of that name does, both legs, the split branch, the pull request and the topic on that repository #499 lines;
- the .github -> ci and hash-object lines;
- the tracking_branch line.
  Each original's ruling is its copy's.
The originals of L4-09, L4-10, L4-44 and L4-46 are FIXED, as their ticked copies say.

Lane: openreposhape-3 (openRepoShape-3)

adopt-project.py, shape_materialize.py: five printed sentences say wh…

adopt-project.py: `plan` asks the leg-path, pin, date and value refus…

adopt-project.py: `check` names a symlink target with a NUL after its…

adopt-project.py: `check` and `execute` ask a plan's `org:` and `id:`…

scaffold-project.py, adopt-project.py and shape_materialize.py: one f…

adopt-project.py: the split commit's message and `plan`'s report name…

adopt-project.py: plan and check take an empty flag or a bare plan va…

(lane openRepoShape-3, 2026-10-20T03:50:13Z) Left over from #902 (PR #1016): in adopt-project.py, elected_by:, reference: and org: in a hand-edited adoption plan, read by the plan's reader as a number or a boolean (elected_by: 007, elected_by: true, elected_by: 1.50, reference: 1.50, reference: 007, org: 007, org: true), still give check exit 0 and plan ok, and execute takes str() of what was read (007 is 7, true is True, 1.50 is 1.5). It is the same class as L1-K3, which #1016 closes for id:, tracking_branch: and adopt_branch: only. Measured on main c4dd19d and on #1016's head 9261ff2. Each key needs a finding of its own: org: is read as str on purpose for the not-the-source comparison (#513's review), visibility: true is already the validator's finding, and an elected_on: read as a number is #294's. Direction: refuse each by what was read, under a code of its own, as #1016 does for the three keys, with NOT_TEXT_REMEDIATION's remedy.

Lane: openreposhape-3 (openRepoShape-3)

adopt-project.py: `check` and `execute` ask a plan's `org:` and `id:`…

adopt-project.py: the split commit's message and `plan`'s report name…

shape_materialize.py, adopt-project.py and bump-leg.py: a tag a clone…

adopt-project.py: `check` names what the clone's `.git` holds as git …

Navigation Menu

Repository navigation

Tracking: adopt-project.py, converting an existing repository in place #457

106 remaining items

Body

Comments, 2026-10-12

Comments, 2026-10-13 and 2026-10-14

Comments, 2026-10-15

Comments, 2026-10-16 and 2026-10-18

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Tracking: adopt-project.py, converting an existing repository in place #457

Description

Activity

106 remaining items

brettheap commented on Oct 18, 2026

brettheap commented on Oct 18, 2026

Body

Comments, 2026-10-12

Comments, 2026-10-13 and 2026-10-14

Comments, 2026-10-15

Comments, 2026-10-16 and 2026-10-18

brettheap commented on Oct 20, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions