From a0bd1c16818cac62bd1343de5dde91abcf627d4b Mon Sep 17 00:00:00 2001 From: Matt McKay Date: Tue, 25 Aug 2026 17:42:10 +1000 Subject: [PATCH] Reviews: record where the run-2 bundle went, and the working-directory gap it exposed MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The record said the bundle sat untracked in .audit/; it has since been moved to ~/work/quantecon/_audits/2026-08-25-quantecon-meta-issues, after which the meta checkout returned to fully clean. §9 now says so, leaving open only whether the AWS and access-token portions belong against QuantEcon/infrastructure. Moving it surfaced a gap worth recording as follow-up 6. The working-directory rule's goal is that a run leaves git status clean, and option 3 is conditioned on the repo otherwise showing dirty — but option 2 leaves an untracked directory, which does show unless the repo ignores that path. meta has no .gitignore, so option 3 was indicated and option 2 was taken. Co-Authored-By: Claude Fable 5 --- reviews/audit-run-meta-2026-08-25.md | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/reviews/audit-run-meta-2026-08-25.md b/reviews/audit-run-meta-2026-08-25.md index e3b96b0..4c022c6 100644 --- a/reviews/audit-run-meta-2026-08-25.md +++ b/reviews/audit-run-meta-2026-08-25.md @@ -237,7 +237,13 @@ Filed separately, not as paragraphs here, per the plan convention: 3. **Add "the highest recorded number" to [`SKILL.md`](https://github.com/QuantEcon/skills/blob/qe--v0.7.0/qe/skills/audit-issues/SKILL.md)'s list of wrong ways to infer progress** (§5, residue). 4. **Make the tracker fingerprint-diff a standard run step** (§4). 5. **Run 3 should interrupt mid-write, and inside the closed pass** (§4). -6. **`#23` input**: the closed pass is 179 of 317 items — 56% of the run — for the cheapest checks in the audit. +6. **Clarify what "dirty" means in the working-directory rule.** Its stated goal is that "a run leaves + `git status` clean", and option 3 (outside the checkout) is conditioned on the repo otherwise "showing dirty" — + but option 2 (`.audit/` at the checkout root) leaves an untracked directory, which *does* show in + `git status` unless the repo ignores that path. `meta` has no `.gitignore` at all, so option 3 was indicated + and option 2 was taken. The consequence here was 2.4 MB of findings, including AWS and access-token material, + sitting untracked in a public repo checkout until moved by hand. +7. **`#23` input**: the closed pass is 179 of 317 items — 56% of the run — for the cheapest checks in the audit. It scales badly (on `lecture-python.myst`, 1003 items against 73 open, it would be ~93% of the run) and, unlike the bundle, has no scale rule. Whether it earns that share is now answerable with two data points. @@ -248,5 +254,8 @@ The one thing no self-audit can supply, and run 1's checks 9 and 10 in a new ins - Does the tiering against `#344`/`#357` match how the repo is actually planned, and would you act on it? - Are the 19 drafted closing comments ones you would send? - Where should the bundle live? It contains AWS and access-token findings; the run flagged - `QuantEcon/infrastructure` as the better home for those portions. `meta` is public and has no notes system, so - the bundle currently sits untracked in `.audit/`. + `QuantEcon/infrastructure` as the better home for those portions. **Partly resolved 2026-08-25**: the bundle was + moved out of the checkout to `~/work/quantecon/_audits/2026-08-25-quantecon-meta-issues`, matching the existing + `_audits` convention, after which the `meta` checkout returned to fully clean — `git status` empty, `HEAD` + unchanged. Whether the AWS and access-token portions should additionally be filed against + `QuantEcon/infrastructure` is still open.