Gate the pattern-detectable half of the representative-data rule - #596
Conversation
Selection step 9 applied to c8687c5. The cluster it carried filed no issue, so nothing moves to "Verified Complete, Awaiting Close" and the deletion is the whole of the bookkeeping for it. "The Representative-Data Path Check" waited on exactly that floor, on the reasoning that gating a path shape is worth having only once the gate can prove it read something, so its state goes to ready and the dependency is recorded as settled rather than dropped silently. Its freshness anchor is rewritten against the tree rather than against the issue, per step 4: the rule is stated in GOVERNANCE.md and none of the gate's nine rules reads a path shape. Its one open question survives untouched, whether a home path in an operational repo's runbook is a finding, since that is a scoping decision with fleet reach rather than something this commit settles. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
GOVERNANCE.md states that agent-authored text illustrates with constructed data rather than data observed in the maintainer's environment, and says in the same breath that no pattern closes it. This gates the half a pattern can reach and is introduced as that half, since a check mistaken for the answer is what stops anyone looking at the name-shaped exposure the rule exists for. The shapes were chosen by measurement rather than from the list the backlog recorded. A bare drive letter is not one of them: against this repository it matched 11 files and named a path in none, because an escaped newline after any word ending in a letter and a colon reads as a drive letter, so a YAML fixture is full of them. The shape kept is a drive letter followed by Users. The exemption carries the burden here, so it is built rather than listed. A real user segment is required, which is what lets the rule's own wording quote its shapes without matching them, so no exemption names the files that describe the rule and none goes stale as they move. Container and runner accounts are exempt by name because every home path in this repository, all 15, is /home/vscode from the devcontainer snippets and their doc: without that exemption the rule would open with 15 findings and no true positive. An operational repository is exempt entirely, since its runbook carries the literal path an operator types, which is the repository's own content rather than an agent quoting an environment. The model is read from what the repository carries, spec/files.json declaring the operational payload for one model and the release payload for the other, so nothing reaches the hub registry to ask and the hub, carrying both, reads as release rather than exempting itself from a rule it authors. The skip is announced, because a rule that silently stops running reads as a rule that passed. Unlike every other prose rule this one reads fenced blocks and config values, since a transcript pasted from a terminal and a bind mount naming a real home are the exposure in its most consequential forms rather than prose. Both source files are read by the rule they implement, so the bait paths are assembled from parts and a case asserts neither file carries a literal. That is the approach this file already takes for non-ASCII, which it writes as escapes for the same reason. The rule reports nothing across the tracked tree and leaves the #519 backlog at 557 violations across 45 files, unchanged. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
Pull request overview
Adds a new home-path rule to scripts/prose_lint.py to gate the pattern-detectable subset of the "Representative Data in Agent-Authored Text" policy (absolute home paths that look like real user homes), including tests and an operational-repository exemption, and updates the corresponding TODO cluster state.
Changes:
- Add
home-pathdetection toscripts/prose_lint.py, include it inDEFAULT_RULES, and exempt operational checkouts (with an explicit stderr note). - Add focused unit tests for
home-path, including placeholder non-matches, service-account exemptions, fenced/config coverage, and operational exemption behavior. - Update
TODO.mdto mark the representative-data path check cluster asreadyand refresh itsCheckedanchor.
Reviewed changes
Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.
| File | Description |
|---|---|
| TODO.md | Updates the work-cluster state/anchors for the representative-data path check. |
| scripts/prose_lint.py | Implements the home-path rule, adds it to defaults, and adds an operational-checkout exemption. |
| scripts/test_prose_lint.py | Adds unit tests covering the new home-path rule and the operational exemption behavior. |
Suppressed comments (1)
TODO.md:37
- This list of shapes still includes "a bare drive letter", but the new home-path regex intentionally excludes that form due to false positives. Consider updating the shapes here to match what the gate actually checks.
- **Settled** - The shapes are `/home/<name>`, `/Users/<name>`, `C:\Users\<name>`, and a bare drive letter.
The review found the entry still promising a bare drive letter after the implementation dropped that shape, which is a contradiction between the backlog and the tree. Correcting the bullet is the smaller half of the answer: this branch is the pull request that ships the cluster, so step 9 deletes it rather than leaving a corrected description of finished work. The claim the corpus disproved is recorded where a reader will find it, in this branch's commit message and pull request description, which is where the measurement that disproved it also lives. A bare drive letter matched 11 files here and named a path in none. The orphaned prose-lint link definition goes with it, since MD053 is enabled by default and the two clusters that cited it have both shipped. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 3 out of 3 changed files in this pull request and generated no new comments.
Suppressed comments (3)
scripts/prose_lint.py:64
- HOME_PATH matches only "\Users\" with a capital U. Windows paths are case-insensitive, so a pasted path like "C:\users\alice" would currently be missed (false negative). Consider allowing lowercase as well to match real-world logs and terminal output.
HOME_PATH = re.compile(r'(?:/home/|/Users/|[A-Za-z]:\\Users\\)(?P<user>[A-Za-z][A-Za-z0-9._-]*)')
scripts/prose_lint.py:160
- home_path_findings includes the full matched absolute path in the lint output. Since the purpose of this rule is to prevent leaking real home paths, echoing the literal path into CI logs can unnecessarily amplify the exposure. The file+line is already reported, so the message can stay actionable without repeating the sensitive value.
out.append((lineno, 'home-path',
f'absolute home path {m.group(0)!r} -> use a constructed path, not an '
'observed one'))
scripts/test_prose_lint.py:1585
- Test name reads as a grammatical typo ("persons" vs "person"). Renaming improves readability and avoids introducing a new inconsistent identifier.
def test_a_container_account_is_not_a_persons_home(self) -> None:
The review found a false negative: the Windows branch required a literal Users while that filesystem is case-insensitive, so C:\users\alice and c:\USERS\alice were both missed. A pasted terminal path is the case this rule exists for and it carries whatever casing the shell printed. The fix is scoped to the drive-letter branch rather than applied to the whole pattern, which is the part worth stating. A lowercase /users/ is the commonest REST path there is, so widening the POSIX branch the same way would flag an API route in every document that has one. Two cases hold both halves. The container-account test loses a possessive that read as a typo. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
Answering the three suppressed comments from the round on Accepted,
The fix is scoped to the drive-letter branch rather than applied to the whole pattern, which is the part worth recording. Widening the POSIX branch the same way would match Accepted, Declined, The case the finding is reaching for is a path that never reaches a commit, which this gate cannot see, since it reads committed files. That gap is genuine and is stated in the description rather than closed here: the leak that motivated the rule was in a pull request comment, which no committed-file linter reads. |
… that blocks on what it cannot read (#609) Promotes 14 commits from `develop`. Merge with a **merge commit** (`gh pr merge --merge`), never a squash, and **without `--delete-branch`**, since this pull request's head is `develop` itself. Closes #607 through the closing keyword already carried in `530cf71`, which is why it is not repeated here. ## The prose backlog, cleared end to end `#600`, `#604`, `#605`, `#606` took the tree from **557 findings across 45 files to 0 across 0**, in four batches ordered by surface: snippets, comments, hub-only Markdown, then the carried files. Each batch measured the checker's own exemption against the live corpus *before* sweeping, and twice the measured answer was **"do not change the checker"**, which is a result of that pass rather than a skipped one. `#594` added the floor that makes those numbers trustworthy: a diff-scoped run now asserts what it actually scanned, since a check whose scan matches nothing reports zero findings and reads exactly like a pass. One finding from that work is worth carrying up: an exemption that is too **loose** produces silence rather than false positives. #519 recorded the governance files as clean; today's checker reports 38 findings against those same files as they stood at the commit that measured them. ## A review loop that fails closed `#599`, `#601`, `#602`, `#603` and `#608` are one arc on `scripts/pr_review.py`, each removing a shape in which the loop reported a clean pass over a review it had misread: - **`#599`** removed the shape a reply kept failing in, by taking the thread's *words* rather than an id, so there is no argument a hand-typed `PRRT_...` fits in. - **`#602`** made `claims` resolve what a description points at rather than what it looks like. - **`#603`** gave a disproved claim a home the next round reads. - **`#608`** reads the file-coverage line, and then generalizes: every reader keys on a structural marker, so a marker that changes spelling is a section the reader stops finding and reports as absent. The digest now vets headings, `<summary>` texts, metadata labels, coverage wordings and the reviewer login against an inventory measured from **332 review bodies**, and **blocks on anything outside it**, exit `43`, with the remedy stated as filing an issue on the hub. Whether to merge regardless is the maintainer's decision. `GOVERNANCE.md` merge gate went from four preconditions to **five** accordingly. ## Governance and tooling - **`#593`** states which checkout an agent works in and what the hub is, which is the host-wide routing the repositories that most need it cannot carry. - **`#596`** gates the pattern-detectable half of the representative-data rule, honest that no pattern closes the name-shaped case. - **`#598`** declares where a repository states what CI cannot verify. - **`#592`** regrouped `TODO.md` by what ships rather than by what it touches, so a `###` heading is one pull request. - **`#601`** ended a `gh push` argument list at a newline rather than only at `&&`, fixing a write-guard over-block. ## Verification Run on `develop` at `530cf71` immediately before opening this: `test_pr_review.py` (174), `test_prose_lint.py`, `test_repo_gate.py`, `spec/audit.py --selftest`, `gh-write-guard.py --selftest`, `spec/validate.py`, `repo_gate.py`, the prose gate in both CI invocations, markdownlint and editorconfig-checker. All clean. ## Not carried by this promotion - **#519 is complete and still open.** `TODO.md` holds its closing evidence under "Verified Complete, Awaiting Close". Closing it is the maintainer's call, so no keyword for it appears here. - **The re-vendor debt is now nine files.** `#606` queued seven, and `#608` changed `GOVERNANCE.md` "PR Review Etiquette" (`verbatim`) and `.github/copilot-instructions.md` (`intent`) on top. The `intent` half produces no hash and therefore no audit finding, which is why the Fleet Sweeps entry names those files by hand. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
Why
GOVERNANCE.md"Representative Data in Agent-Authored Text" says agent-authored text illustrates with constructed data rather than data observed in the maintainer's environment, and says in the same breath that no pattern closes it. This gates the half a pattern can reach.It is introduced as that half deliberately. The leak that motivated the rule was name-shaped and sat in a pull request comment, which no committed-file linter reads, so a check mistaken for the answer is what stops anyone looking at the exposure the rule actually exists for.
Carries the
TODO.mdcluster "The Representative-Data Path Check", which #595 released fromblockedonce the scope floor shipped in #594.The shapes were measured, not assumed
The backlog recorded four shapes,
/home/<name>,/Users/<name>,C:\Users\<name>, and a bare drive letter. Run against this repository, a bare drive letter matched 11 files and named a path in none of them.An escaped newline after any word ending in a letter and a colon reads as a drive letter, so a YAML fixture is full of them:
e:\n needsjobs:\n needsinside a test fixture strings:\n ajobs:\n a:inside a test fixture stringC:\tmp(8 hits)That shape is dropped. The one kept is a drive letter followed by
Users, which matches nothing in the tree today. This is narrower than the committed prose, which says "an absolute home path or a drive letter", so the narrowing is stated here rather than left for a reader to discover.The exemption is built, not listed
A wrong exemption hands out a work list that damages correct documents, and this rule has two self-referential problems that a file-name allowlist would solve badly and briefly.
The rule's own wording quotes the shapes it forbids. Requiring a real user segment solves it by construction: the documentation placeholder is not a valid account name, so it describes the shape without matching it. No exemption names
GOVERNANCE.md, and none goes stale when the text moves.C:\Users\<name>in documentation/home/<name>in documentation/home/vscode/.sshContainer accounts are exempt by name. Every home path in this repository, all 15, is
/home/vscodefrom the devcontainer snippets and the doc describing them. Without that exemption the rule would open with 15 findings and no true positive among them.The gate reads its own source, so the bait paths in the tests are assembled from parts and a case asserts that neither
prose_lint.pynor its test file carries a literal. That is the approach the file already takes for non-ASCII, which it writes as escapes for exactly this reason.Operational repositories are exempt, and told so
An operational repository's runbook carries the literal path an operator types. That is the repository's own content rather than an agent quoting an environment it observed, which is the distinction the rule is about.
The model is read from what the repository carries rather than from the hub registry, which a downstream checkout cannot see:
spec/files.jsondeclaresrepo-config/operational/develop.jsonfor the operational model andrepo-config/develop.jsonfor the release one. The hub carries both, being the template for each, so it reads as release rather than exempting itself from a rule it authors.The skip is announced on stderr, because a rule that silently stops running reads as a rule that passed, which is the same failure #594 closed on the other side.
Scope note
Unlike every other prose rule, this one reads fenced blocks and configuration values rather than stripped prose. A transcript pasted from a terminal and a bind mount naming a real home are the exposure in its most consequential forms, and neither is a sentence.
Verification
python3 scripts/test_prose_lint.py- 188 pass, up from 173. Of the 14 new cases, 10 fail againstdevelopand 4 pass against both, holding the must-not-fire direction.python3 scripts/prose_lint.py . --check home-path- clean across the tracked tree.python3 scripts/prose_lint.py . --summary- 557 violations across 45 files, the prose_lint: the hub's own docs, spec, and catalog do not pass the gate it ships #519 backlog figure unchanged, withhome-pathcontributing none. Adding a rule toDEFAULT_RULESdid not move the number.python3 spec/validate.py,python3 spec/audit.py --selftest,python3 scripts/repo_gate.py- clean.Correction: this branch carries the TODO bookkeeping too
An earlier revision of this description said the
TODO.mdedits were held back to avoid conflicting with #595. That was wrong. This branch was cut from #595's branch rather than fromdevelop, so it already contained that commit, and the description claimed an absence a reader could disprove from the diff.The review then caught the real consequence: the cluster entry still promised a bare drive letter after the implementation dropped that shape. Correcting the bullet was the smaller half of the answer, since this is the pull request that ships the cluster, so
c6fca9eapplies step 9 and deletes it rather than leaving a corrected description of finished work. The disproved claim is recorded here and in the commit message, alongside the measurement that disproved it.This makes #595 redundant, since every line of it is already in this branch. Closing it as superseded is the maintainer's call rather than something this branch does.
🤖 Generated with Claude Code