chore(gates,docs): align ruff hook with CI and remove the broken AWS hook - #28
Open
Yosoyepa wants to merge 1 commit into
Open
chore(gates,docs): align ruff hook with CI and remove the broken AWS hook#28Yosoyepa wants to merge 1 commit into
Yosoyepa wants to merge 1 commit into
Conversation
Two audit-visible inconsistencies found while scoping phase 12. 1. `.pre-commit-config.yaml` pinned `ruff-pre-commit` at `v0.11.0` while `pyproject.toml` requires `ruff>=0.16.1` (locked at 0.16.1). Both read the same phase-10 `select` list, but 0.11.0 did not implement 30 of the rules those families now cover, 13 of them stable rather than preview-gated. The local gate was therefore strictly weaker than `uv run ruff check .` in CI. Verified by probe: a `value in []` body passes the old hook and trips RUF060 under the new one. Also moved to the canonical `ruff-check` id, since the bare `ruff` id is now published as a legacy alias. `detect-aws-credentials` is removed outright. The project has no AWS surface -- no SDK dependency, no deployment target, no credential to leak -- so the hook could only ever fire on a false positive, and with no credentials file present it exited nonzero on their absence rather than on a finding, failing the whole `pre-commit run --all-files` gate for a non-finding. This retires the R-04 risk phase 09 accepted. Repository secret scanning stays with gitleaks in the required `security` job, which covers `AKIA`/`ASIA` key ids among every other pattern, so this drops a broken check rather than a layer of protection. `maintainer-workflow.md` no longer claims the hook exists. 2. Phase logs 05, 07 and 08 still carried `PULL REQUEST OPEN` / `IN_PROGRESS` headers with `<pendiente>` PR and merge fields although their PRs are merged; phase 07 also left wave-1 slots at `PENDING` and phase 09 left A4/A5 `IN_PROGRESS` with an unrecorded rollback rehearsal. Every PR number, merge SHA and slot commit recorded here was read back from git history, not inferred. Phase 06's `Release` becomes `UNRELEASED` rather than a tag it does not belong to, and phase 05's prerelease is explicitly left unclaimed because it could not be verified from this checkout. The phase-09 evidence table keeps its original `PARCIAL` record: it describes what that run saw. No source, test, eval case or public contract changed. Gates: uv lock --check, ruff check, mypy (118 files), compileall, uv build, pytest against PostgreSQL 16.14 (853 passed, 3 allowlisted skips, 58 subtests), coverage 92% (threshold 85), L1 eval suite 299/299, pip-audit, pre-commit (4/4, all green), git diff --check. diff-cover reports no coverage-bearing lines in the diff, as expected for a config-and-docs change. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Two audit-visible inconsistencies found while scoping phase 12. No source, test, eval case or public contract changed.
1. The local gate was weaker than CI, and one hook could only fail
.pre-commit-config.yamlpinnedruff-pre-commitatv0.11.0whilepyproject.tomlrequiresruff>=0.16.1(locked at 0.16.1). Both read the same phase-10selectlist, but 0.11.0 did not implement 30 of the rules those families now cover — 13 of them stable rather than preview-gated (B912, ISC004, PLR1708, UP050, RUF060/061/063/064/068/076, RUF102/103/104).Verified by probe rather than by reading changelogs: a
value in []body passes the old hook and trips RUF060 under the new one. Moved to the canonicalruff-checkid, since the bareruffid is now published as a legacy alias.detect-aws-credentialsis removed outright. The project has no AWS surface — no SDK dependency, no deployment target, no credential to leak — so the hook could only ever fire on a false positive. With no credentials file present it exited nonzero on their absence rather than on a finding, failing the wholepre-commit run --all-filesgate for a non-finding. This retires the R-04 risk phase 09 accepted.Repository secret scanning stays with
gitleaksin the requiredsecurityjob, which coversAKIA/ASIAkey ids among every other pattern — so this drops a broken check, not a layer of protection.maintainer-workflow.mdno longer claims the hook exists.2. Stale phase logs
Logs 05, 07 and 08 still carried
PULL REQUEST OPEN/IN_PROGRESSheaders with<pendiente>PR and merge fields although their PRs are merged; 07 also left wave-1 slots atPENDING, and 09 left A4/A5IN_PROGRESSwith an unrecorded rollback rehearsal.Every PR number, merge SHA and slot commit recorded here was read back from git history, not inferred. Phase 06's
ReleasebecomesUNRELEASEDrather than a tag it does not belong to, and phase 05's prerelease is left explicitly unclaimed because it could not be verified from this checkout. The phase-09 evidence table keeps its originalPARCIALrecord: it describes what that run actually saw.Gates
uv lock --check(76 packages, no change) ·ruff check .·mypy src(118 files) ·pytestagainst PostgreSQL 16 — 853 passed, 3 allowlisted skips, 58 subtests · coverage 92% (threshold 85) · L1 eval suite 299/299 ·compileall·uv build·pip-audit·pre-commit run --all-files4/4 all green ·git diff --check.diff-cover reports no coverage-bearing lines in the diff, as expected for a config-and-docs change.
🤖 Generated with Claude Code