[brettheap]

Lane: openreposhape-1 (openRepoShape-1)

Closes #927, from lane 3's triage of #459 (U074, L2-70).

What

Every remedy that tells someone to run family.py bump now names --family-root, so each one runs as printed. bump requires --family-root (the usage at scripts/family.py:7).

• scripts/family.py: the member-already-present and member-duplicate-id remedies name --family-root <holder>. The holder is quoted with shell_quoted(family.root), as add's own line is.
• scripts/bump-leg.py: the bump-leg-assembly-is-not-a-leg remedy names --family-root <holder>, a placeholder, because bump-leg has no holder to name.
• Two more in bump-leg, in their own commit (7b42c652): bump-leg-root-not-a-project and bump-leg-wrong-kind. Each pointed at a bare scripts/family.py bump, and now gives the full line.
• README and the handbook (3df537a4): README's "Windows, natively" sentence and the handbook's Windows panel now name family.py's bump line among the lines quoted for the shell they are printed in, because the two remedies above quote the holder for it. tests/test_readme_windows_quoting_sentence.py holds both to the code; CI's full suite caught the gap (run 39311987496).
• The sweep: git grep and a multi-line scan found 36 hits. Every other hit already carries --family-root, or is prose or history with no line to type. No pinned copy prints a bare bump.

Tests

All runs used /usr/bin/python3 -m pytest under a stub gh, a bogus GH_TOKEN, an empty GH_CONFIG_DIR, GIT_CONFIG_GLOBAL=/dev/null and GIT_SSH_COMMAND=false.

• tests/test_bump_remedies_family_root.py, 21 rows:
  • family.py: both codes, in three holder folders (a plain one, my work and Dana's work). Each checks three things:
    • the printed line, and that shlex gives the holder back as one word;
    • that family.py's own parser accepts it;
    • that it runs for real as printed: exit 0, the pin moved, the holder clean.
  • bump-leg.py: the three remedies, each also accepted by family.py's parser.
• Against cf66102f's code: all 21 fail. The 12 parses and runs_as_printed rows fail on family.py bump: error: the following arguments are required: --family-root; the 6 carries_the_holder_whole rows and the bump-leg-assembly-is-not-a-leg row fail on the printed line, which has no --family-root; and the bump-leg-root-not-a-project and bump-leg-wrong-kind rows find no bump line at all, since both remedies named a bare scripts/family.py bump.
• tests/test_family.py and tests/test_family_add_member_id.py follow the new words. The latter's bump row now runs the printed line exactly.
• Fifteen family, bump-leg and docs files: 585 passed and 1 skipped, which needs a case-insensitive disk.
• The review's fix round (1baafc95): the printed bump line is split by the host's own words, and a filled-in holder is quoted, so a holder path with a space is one argument. Under a basetemp with a space, 4 rows failed at 7b42c652 and pass at the head.
• At 3df537a4: test_readme_windows_quoting_sentence.py 44, test_repo_hygiene.py 100, test_readme_sections.py 76, test_setup_by_hand_commands.py 5 and test_bump_remedies_family_root.py 21 passed.

Notes

• Merges: clean with main 3864b18c, and with shape-doctor.py and shape_advisory.py: a path in the root this user cannot reach is named on every Python, never read as absent or stood in for by the template, a root it cannot search is refused first, and a row that could not ask is never DRIFTED #975, the other open PR that changes README and the handbook, in another section.
• Follow-up, outside bump: family.py: the member-duplicate-id and member-unknown remedies name family.py remove and family.py add without --family-root, so each fails as printed #1005. member-duplicate-id also says `family.py remove --member <P>`, and bump's member-unknown says family.py add; both verbs require --family-root too.
• Squash message: the head's commits include two WIP (#927) ones, and the landing squash carries its own message.
• Review: lane 1's independent reviewer found 3df537a4 READY (family.py and bump-leg.py: every remedy that names family.py bump names --family-root, so each runs as printed (#927) #1002 issuecomment-6278129416).

🤖 Generated with Claude Code

Summary by Sourcery

Make all printed family.py bump remediations executable by supplying and correctly quoting the required family root.

Bug Fixes:

• Ensure every remediation that recommends family.py bump includes the required --family-root argument so the printed command can be executed successfully.

Enhancements:

• Quote family holder paths in generated bump commands and use a placeholder where bump-leg cannot determine the holder.
• Keep the README and handbook Windows shell-quoting guidance aligned with the expanded set of quoted bump commands.

Documentation:

• Update README and handbook guidance to include family.py bump among commands quoted for Windows shells.

Tests:

• Add coverage verifying bump remediation commands preserve holder paths, parse correctly, and run as printed, including paths containing spaces and apostrophes.

[sourcery-ai]

Reviewer's Guide

The PR fixes five remediation messages that printed incomplete family.py bump commands: family-aware errors now include a shell-quoted --family-root path, while bump-leg errors use a <holder> placeholder. Tests verify the commands parse correctly, preserve paths containing spaces or apostrophes, and—where a concrete holder is available—run as printed and update the pin.

Sequence diagram for validated family bump remediation

sequenceDiagram
    participant AddCheck as family.py add checks
    participant Holder as Family holder
    participant CLI as family.py bump
    participant Pin as Member pin
    AddCheck->>Holder: shell_quoted(family.root)
    AddCheck-->>CLI: Print family.py bump --family-root <holder> --member <Project> --to <commit>
    CLI->>CLI: Parse remediation command
    CLI->>Pin: Move member pin
    Pin-->>CLI: Exit 0

Loading

Flow diagram for bump-leg remediation commands

flowchart TD
    A[bump-leg validation] --> B{Remediation reason}
    B -->|assembly is not a leg| C[Print family.py bump --family-root <holder> --member <Project> --to <sha>]
    B -->|root is not a project| D[Print family.py bump --family-root <holder> --member <Project> --to <sha>]
    B -->|wrong kind| D
    C --> E[family.py parser accepts command]
    D --> E

Loading

File-Level Changes

Change	Details	Files
Make every printed family.py bump remediation executable by supplying the required family-root argument.	Add a shell-quoted holder path to the two family add refusal remedies that know the holder. Use a <holder> placeholder in all three bump-leg remedies, including the two newly corrected bare commands. Update affected tests and shared expected remediation text; add coverage for parsing, shell-safe quoting, and real execution.	scripts/family.py scripts/bump-leg.py tests/test_bump_remedies_family_root.py tests/test_family.py tests/test_family_add_member_id.py

Assessment against linked issues

Issue	Objective	Addressed	Explanation
#927	Update the member-already-present and member-duplicate-id remedies in scripts/family.py to include the required --family-root and identify the holder correctly.	✅
#927	Update every scripts/bump-leg.py remedy that names family.py bump to include --family-root, using a holder placeholder where the holder is unknown.	✅

Possibly linked issues

• #U074: The PR fixes all three missing --family-root arguments identified by the issue, making those remedies runnable as printed.

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[brettheap]

(lane openRepoShape-1, 2026-10-19T16:06:38Z) Independent review of #1002 (#927) by r1002, PARTIAL at 7b42c652 (stopped at lane 1's usage checkpoint): NOT READY, provisional. The code works: every changed remedy ran as printed, the sweep is complete, and every merge is clean. The one MUST was a false sentence in the body, now replaced with r1002's exact words. The two SHOULDs (a holder path with a space, in two test rows) and the NITs go into a fix round after the shared window resets (20:30Z). r1002 then finishes the checks it lists.

r1002 report at checkpoint: NOT READY at 7b42c652 (provisional)

The PR's code works. Every changed remedy ran for real as printed, the sweep is complete and every merge is clean. The verdict is NOT READY because of one false sentence in the PR body, which is a MUST under the brief. Fixing it means editing the PR body only; the head does not change. I stopped at your checkpoint, so some checks are left (listed at the end). The worktree is clean: git status --short prints nothing at HEAD 7b42c652. Mutant M4 was half-applied when I stopped; I put it back with git checkout -- scripts/family.py. No test or mutant process is still running.

Findings, most severe first

MUST: a false sentence in the PR body (pr927-body.md:24)

• The body says: "Against cf66102f's code: all 21 fail with 'the following arguments are required: --family-root'".
• I put cf66102f's scripts/family.py and scripts/bump-leg.py in place and ran the new file: 21 failed. Only 12 of them fail on that message (log logs/old-new-file.log):
  • 6 parses rows: family.py bump: error: the following arguments are required: --family-root.
  • 6 runs_as_printed rows: assert 'the followi...are required' not in ....
  • 6 carries_the_holder_whole rows: assert 'bump --membe...--to <commit>' == 'bump --famil...--to <commit>'.
  • The assembly-is-not-a-leg row fails the same printed-line comparison.
  • The root-not-a-project and wrong-kind rows fail assert 0 == 1 in the_bump_line, because no bump line is printed at all.
• Fix, in these words: "Against cf66102f's code: all 21 fail. The 12 parses and runs_as_printed rows fail on family.py bump: error: the following arguments are required: --family-root; the 6 carries_the_holder_whole rows and the bump-leg-assembly-is-not-a-leg row fail on the printed line, which has no --family-root; and the bump-leg-root-not-a-project and bump-leg-wrong-kind rows find no bump line at all, since both remedies named a bare scripts/family.py bump."

SHOULD: a changed test row now breaks the holder path it reads (tests/test_family_add_member_id.py:177)

• The line is argv = commands[0].replace("<commit>", older).split(). The printed bump line now carries the holder, quoted, and .split() cuts it at a space. Before this PR the row passed str(holder) as one argument.
• With --basetemp=.../r1002/bt/my work at the head: [bump] FAILED with family.py: error: unrecognized arguments: work/test_each_command_the_refusal_0/InkRouter', and [remove] passed.
• At cf66102f's code and test, same basetemp: both rows passed.
• With my fix applied (then restored): both rows passed.
• Fix: argv = host_words(commands[0].replace("<commit>", older)), and add from test_family_printed_quoting import host_words.

SHOULD, suspected and not run: tests/test_bump_remedies_family_root.py:220 puts an unquoted str(project.parent) in for <holder> before host_words, so a basetemp holding a space should break it the same way. If confirmed, the fix is .replace("<holder>", shell_quoted(project.parent)).

NIT: the new test's docstring is stale (tests/test_bump_remedies_family_root.py:5-7)

• It says "Three remedies ... each ended in ...: --family-root", then lists five codes.
• The two remedies fixed in 7b42c652 named a bare scripts/family.py bump. Typed as printed, argparse should also list --member and --to; I have not run that.
• Fix: "Five remedies named the verb without it. Typed as printed, the three that named --member and --to ended in ...: --family-root, and the two that named a bare scripts/family.py bump in ...: --family-root, --member, --to."
• Lines 18-19 also say "two with a space and an apostrophe"; only Dana's work has an apostrophe.

NIT: one placeholder where the path is known (scripts/bump-leg.py:215-223)

• When --root holds a family.yaml (the test's a_holder case), the run could name that root as --family-root instead of <holder>.
• Not blocking: <Project> and <sha> stay placeholders anyway.

Item 3: the remove line (confirmed; belongs in a follow-up)

• Typed as printed in all three holders, remove --member IRRS exits 2 with family.py remove: error: the following arguments are required: --family-root.
• Two more remedies outside bump have the same gap:
  • scripts/family.py:8065: bump's own member-duplicate-id names family.py remove --member {held['project']}.
  • scripts/family.py:8721: bump's member-unknown names family.py add --member <org>/<P>.
• The remedy for item 3 is at scripts/family.py:6297.
• I recommend one follow-up issue covering all three, not this PR: scripts/family.py and scripts/bump-leg.py: three remedies name family.py bump --member ... --to ... without the --family-root that bump requires #927's scope is bump, and fixing one of the three here leaves that sweep half done.
• gh issue list --search "family-root in:title" found no existing issue for these.

Item 1: the sweep

• I scanned joined string literals with ast across scripts/, templates/, the root tools, src/, AGENTS.md, README.md and docs/. Every typed family.py bump line carries --family-root:
  • family.py:6179-6180 and 6298-6299;
  • bump-leg.py:219, 233 and 288;
  • the template validate-family.py:207;
  • AGENTS.md:376, README.md:954, handbook.html:1417, the two templates/family-root docs and docs/cli.md:979.
• Every other hit is prose or history.
• No pinned copy changed: neither scripts/family.py nor scripts/bump-leg.py is in any copy list in shape_materialize.py.
• The quoting matches _add_command (host dialect), and the holder printed is the resolved family.root from Family.open at all three call sites.

Item 2: real runs

The probe (probe/test_r1002_probe.py) passed all 7 rows:

• Family remedies: both codes, with three holders:
  • a holder under Dana's my work;
  • a relative --family-root Dana's work/InkRouter;
  • a holder named through a symlink (link's place).
• How they were run: each printed line was typed into bash, sh and zsh from another folder. Every run exited 0, the pin moved, the holder stayed clean and validate-family passed.
• Paths printed: the relative and symlinked holders are printed as their absolute, resolved paths.
• bump-leg remedies: all three, with the placeholders filled in, typed in bash from the standard's checkout. Each exited 0 and moved the pin.

Mutants (3 of 13 run before the checkpoint)

Mutant	Result	Evidence
M1 member-already-present without shell_quoted	killed	new file: 6 failed
M2 member-duplicate-id without shell_quoted	killed	new file: 6 failed
M3 the holder's parent folder named instead of the holder	killed	new file: 9 failed; test_family: 4 failed
M4 member-duplicate-id site reverted	interrupted	restored, no result

Test counts at 7b42c652

• New file 21; test_bump_leg 54 + 1 skipped (the case-insensitive-disk row); test_family 112; test_family_add_member_id 14; test_refusal_default_remediation 20; test_repo_hygiene 100; test_cli_reference 23; test_agents_md_codes 47.
• Total: 391 passed, 1 skipped.
• At cf66102f's code: new file 21 failed; test_family_add_member_id 5 failed and 9 passed; test_family -k already_pins 4 failed.

Merges

• git merge-tree --write-tree of 7b42c652 with each of these is clean (exit 0):
  • main 2a78ae1a;
  • family.py: bump's fetch no longer moves the mount's own branches, works where the mount maps tags into origin/ instead of telling the person to fix the network, reads its refspecs with git config --list -z, and asks git remote get-url origin first, so a mount with no origin is not fetched from a path of that name (#905) #946 5d5d06bb, family.py: a holder, a member or its object store this user cannot read is refused by name, exit 2, on every Python, never family-root-missing, member-uninitialized or "not on disk", and no rollback removes past it (#430) #955 0606cf92, family.py add: a commit that records the member's gitlink anywhere but members/<Project> is taken back and refused, naming the path git chose, and members: null, Null, NULL or ~ takes the first member as members: does (#920) #957 d673e090, family.py init: where the holder's own YAML reader refuses the manifest it would write, family-manifest-invalid names the value on the line it refuses, --org is '@x', and gives the reader's remedy, not the validator's (#929) #960 dc5277e6;
  • bump-leg.py and family.py: the tag check compares the commits tags point to, so a lightweight or another annotated tag at the same commit reaches it (#982) #988 203e9410.
• Each merged tree keeps the same five fixed remedies and adds no bare family.py bump.

Style and the PR text

• git diff --check cf66102f 7b42c652 is clean. No added line is over 79 columns, all added text is ASCII, there are no tabs and no compound asserts.
• The title is true. Every other body sentence I checked is true, except two I could not verify:
  • "36 hits": my single-line grep gives 33, and the total depends on their multi-line scan.
  • "585 passed" across fifteen files: I ran 8 files, not their 15.

Checks left

1. Mutants M4–M13, then a planned M14 (shell_quoted(family.root, "posix"), expected to survive on Linux CI). Resume with mutants.py a M4 M5 M6 M7, then mutants.py b.
2. The new file under a basetemp holding a space, to settle the suspected SHOULD.
3. argparse's message for a bare scripts/family.py bump, for the docstring NIT.
4. tests/test_family_dry_run.py -k "duplicate or already" and tests/test_family_printed_quoting.py.
5. A final git status --short.

All paths below are under /workspace/projects/openRepoShape-worktrees/_lane1-scratch/r1002/. Logs are in logs/ (base-summary.txt, old-new-file.log, probe-head.log, split-head.log, split-base.log, split-fix.log, mutants-a.txt). Scripts are in tools/ (runtests.sh, mutants.py, sweep.py, bumpwords.py).

Lane: openreposhape-1 (openRepoShape-1)

[brettheap]

(lane openRepoShape-1, 2026-10-19T21:33:30Z) Re-check of #1002 (#927) by r1002, at 1baafc95: READY. The fix round (the holder path with a space, and the docstring) is verified. 12 of 14 mutants are killed, and the two survivors are equivalent on the Linux CI host. The merge with main 3864b18c is clean. The follow-up for the remove and add remedies is #1005. #1002 lands at 1baafc95 once its own checks are green, by lane 3's turn rule.

r1002 final report: READY at 1baafc95

All checks on #1002 are done and none is open. The worktree is clean: HEAD is 1baafc95 and git status --short is empty. Every mutant was put back after its run, and nothing is still running. Every pytest run had EMAIL=r1002@openreposhape.invalid set.

Mutants M1 to M14 at 1baafc95

All were run after the reboot. Each was put back with git checkout -- and the tree checked clean. 12 were killed and 2 survived; both survivors are equivalent on the Linux CI host.

Mutant	Result	Failing rows
M1 member-already-present without shell_quoted	killed	new file 6
M2 member-duplicate-id without shell_quoted	killed	new file 6
M3 the holder's parent folder named instead of the holder	killed	new file 9, test_family 4
M4 member-duplicate-id site reverted	killed	new file 9, test_family_add_member_id 5
M5 member-already-present site reverted	killed	new file 9, test_family 4
M6 bump-leg wrong-kind reverted	killed	new file 1
M7 bump-leg root-not-a-project reverted	killed	new file 1
M8 bump-leg assembly-is-not-a-leg reverted	killed	new file 1
M9 placeholder spelled <family-root>	killed	new file 1
M10 shlex.quote(str(root)) at both family sites	killed	new file 2 (carries_the_holder_whole[Dana's work-…])
M11 family.path.parent instead of family.root	survived	equivalent code
M12 a path relative to the folder the command is run from, for member-duplicate-id	killed	new file 6, test_family_add_member_id 4
M13 --family-root printed after --member	killed	new file 3, test_family 4
M14 shell_quoted(family.root, 'posix') at both sites	survived	equivalent on a POSIX host

The two survivors:

• M11 is equivalent code. Family.__init__ sets self.path = root / MANIFEST (scripts/family.py:302), so path.parent is always root. No row can kill it.
• M14 is equivalent on every host the CI runs.
  • On Linux, and under Git Bash, the host dialect is already posix, so the output is the same bytes.
  • On native Windows without MSYSTEM, the existing rows test_the_bump_a_remedy_names_carries_the_holder_whole[Dana's work-member-already-present] and [Dana's work-member-duplicate-id] would kill it. There the expected shell_quoted(placed) is PowerShell's '…Dana''s work…', and the mutant prints '…Dana'\''s work…'.
  • CI is Linux-only. A row that kills it on Linux would have to load family.py in-process, monkeypatch repo_shape.host_shell_dialect to return "nt", and assert that the remedy text holds shell_quoted(root, "nt").
  • No such row exists for _add_command or _switch_command either, so this gap is suite-wide and not this PR's.

Test counts at 1baafc95 (from the earlier round, logs read back)

• The eight brief files in one session: 391 passed and 1 skipped under a basetemp with a space, and again under a plain one. The skip is test_bump_leg.py:697, which needs a disk that ignores case.
• tests/test_family_dry_run.py -k "duplicate or already": 3 passed.
• tests/test_family_printed_quoting.py: 84 passed.
• With 7b42c652's two test files under the spaced basetemp: 4 failed. After the fix, those rows pass.
• A bare scripts/family.py bump prints ...required: --family-root, --member, --to, which matches the mended docstring.

Merge

• I fetched main again; it is now 3864b18c.
• git merge-tree --write-tree 1baafc95 origin/main exits 0 with tree 98770e2c, and there are no conflicts.
• The two commits main gained (41aed9ca, 3864b18c) touch none of the PR's files or the files its tests import.
• In the merged tree all six printed family.py bump lines carry --family-root: bump-leg.py 219, 233 and 288, family.py 6620 and 6749, and the template validate-family.py 220.

Still open, none blocking

• NIT: commit 1baafc95's message says "493 passed" for the eight files; I get 391. The squash carries its own message, so this does not land.
• Follow-up issue: these remedies still print a family.py line without --family-root:
  • scripts/family.py:6297: remove, in add's member-duplicate-id;
  • scripts/family.py:8065: remove, in bump's member-duplicate-id;
  • scripts/family.py:8721: add, in bump's member-unknown.

All files are under /workspace/projects/openRepoShape-worktrees/_lane1-scratch/r1002/: the mutant log is logs/mutants-h.txt, the head-run summary is logs/head-summary.txt, and the mutant driver is tools/mutants.py.

Lane: openreposhape-1 (openRepoShape-1)

[brettheap]

(lane openRepoShape-1, 2026-10-19T21:47:50Z) Re-check of #1002 by r1002, at 3df537a4: READY. CI's full suite failed at 1baafc95 on test_readme_windows_quoting_sentence.py (run 39311987496). 3df537a4 names the bump line in README's Windows sentence and the handbook panel, and in the test's NAMED. #1002 lands at 3df537a4 once its own checks are green.

r1002 on #1002 at 3df537a4: READY

The new commit fixes the one CI failure at 1baafc95, the README sentence is true, the requested tests pass, and every merge is clean. Your list of 8 earlier finding was correct: no new conflict. The worktree is clean and detached at 3df537a4; I put back each file I swapped in temporarily. Every pytest run had EMAIL=r1002@openreposhape.invalid set.

The delta 1baafc95..3df537a4

• What changed: README.md (+2/-2), docs/handbook.html (+12/-12) and tests/test_readme_windows_quoting_sentence.py (+4/-2). No script changed.
• Style: git diff --check is clean, and no added line is over 79 columns. The only non-ASCII added is the § in the seven moved handbook markers, which the file already uses 23 times.
• The CI failure reproduces at 1baafc95: with its README, handbook and test file in place, test_the_functions_that_quote_for_the_host_are_the_ones_the_sentence_names[scripts/family.py] fails with "Left contains 2 more items: {'_checked_add_target': 1, '_checked_member_id': 1}". It passes at 3df537a4.

Is the README sentence true?

Yes.

• family.py prints exactly two bump lines, member-already-present and member-duplicate-id. Both quote the holder with shell_quoted(family.root) in the host dialect, and the earlier sweep found no other printed family.py bump line.
• The test's NAMED now lists those two functions with one quoted use each, which is the count the test reads from the source.
• The --member value is a validated project name, and <commit> is a placeholder. Neither needs quoting in either dialect.
• bump-leg.py's bump lines print a <holder> placeholder, and the template validate-family.py quotes for POSIX. The sentence names neither file, so it claims nothing false about them.

Counts at 3df537a4

246 passed in one session, none failed:

File	Passed
test_readme_windows_quoting_sentence.py	44
test_repo_hygiene.py	100
test_readme_sections.py	76
test_setup_by_hand_commands.py	5
test_bump_remedies_family_root.py	21

The mutant table from the last round (M1 to M14: 12 killed, M11 and M14 equivalent) still holds, because this commit changes no script.

Merges

• git merge-tree --write-tree 3df537a4 origin/main (main 3864b18c) exits 0 with tree 60099e84.

Open PRs that change README.md or docs/handbook.html

gh pr list returned 16 open PRs, each listing at most 8 files, so no file list was cut short. Two touch these files: #1002 itself and #975 (fix/430-c-doctor-unreadable, head 3fd80cbe).

• No conflict. git merge-tree --write-tree 3df537a4 origin/fix/430-c-doctor-unreadable exits 0 with tree 63bd26d2.
• Different sections:
  • family.py and bump-leg.py: every remedy that names family.py bump names --family-root, so each runs as printed (#927) #1002 changes README § "Starting a project in a new organisation" (line 158) and its seven handbook markers.
  • shape-doctor.py and shape_advisory.py: a path in the root this user cannot reach is named on every Python, never read as absent or stood in for by the template, a root it cannot search is refused first, and a row that could not ask is never DRIFTED #975 changes README § "Keeping a project's shape current" (line 1210), two of that section's markers and one handbook sentence.
  • No marker is touched by both.
• Checked in place: with the merged tree's README.md and docs/handbook.html swapped in, test_repo_hygiene.py and test_readme_windows_quoting_sentence.py gave 144 passed. Both PRs' markers stay valid together, in either landing order.

Logs are in /workspace/projects/openRepoShape-worktrees/_lane1-scratch/r1002/logs/: r3-five.log, r3-before.log and r3-merged975.log.

Lane: openreposhape-1 (openRepoShape-1)

[sonarqubecloud]

Quality Gate passed

Issues
1 New issue
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[sourcery-ai]

Sorry @brettheap, you've used your own review budget of 250,000 diff characters for the last 7 days.

You can request another review in 1 day and 16 hours by commenting @sourcery-ai review. Upgrade to get a review now.