Repository navigation
v1.93.4.0 fix(review-ledger): a finding's own Severity field sets its label, not severity words in its text (GST-976) - #3867
Merged
Conversation
…everity words never raise it (GST-976) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
…st and emoji-led labels lead, an inline field floors the next (GST-976 review) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
… space, fields after a parenthesis or without a colon at the line start (GST-976 review) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
…d may be a range; a mid-line Severity field must close its clause (GST-976 review) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
…se; mid-line fields accept : / and → after the label (GST-976 fix-check) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
…kes its strongest label; mid-line fields keep the clause-closing set (GST-976 scoped check) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
…uts, fixtures and generated field shapes; a multi-label Severity field counts as none, lowercase fields keep their label (GST-976) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
|
😎 This pull request was merged. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What changed
/ship's finding collector (
parseFindingsin lib/review-ledger.ts) now reads a finding's severity from its ownSeverityfield. A severity word in the finding's text no longer raises it. Before: in gstack-evals #143 the outside voice marked six findingsBlocks: no, and the "P0" in #1, #2 and #4 (the plan's phase name) collected as P0, so the judge refused the downgrade (JUDGMENT_REJECTED) and the receipt needed accepted_by. After: those three collect at their stated Medium/Low.CRITICAL:,[P0],- High:,1. api.ts:10 CRITICAL:, a label heading) stays the floor.Blocks: yesis still CRITICAL. The native voice's FIXABLE still collects as CRITICAL (normalizer v2; FIXABLE moves to GST-987).test/review-ledger-invariant.test.ts: never weaker than main. Over every input the review-dispatch and prose-voice-floor tests pass, every reader fixture and 1,440 generated field shapes (1,815 findings), no label ranks below main's unless the finding's field states exactly one label, and such a finding collects at that label unless a lead, a label heading, Blocks: yes or native FIXABLE raises it. It fails at b3ddd6c (262 violations) and f3810d4 (31).Why
GST-976 (GSTA-167): a token in a finding's text overrode the reader's verdict and made verify dirty on a docs-only /ship.
Blast radius
lib/review-ledger.ts
parseFindingsonly (collect and verify). Free replay of main vs this head over 80 stored reader outputs (302 findings) and 334 merged lines: 3 findings go from blocking to non-blocking (#143's P0s) and 0 the other way. The same invariant over the stored private outputs: 0 violations over 635 findings.Pins changed
Severity: Highnow collects High (Captain ruling A: the field decides).Severity: Low (not High)now collects Low (Captain ruling A).CRITICAL: x.ts:4 … (pass 1 had Severity: Low)) and :522 (- High: …\nSeverity: Medium) unchanged.Approvals
Review notes (open, below the floor)
LOCUSregex is quadratic on long dash runs (3.4 s at 64,000 characters, same on main). Not fixed here._or1.runs cost 142–373 ms in the lead regex (less thanLOCUSon the same lines).x.ts:3 – 5 CRITICAL:) still hides a lead label (same as before).Severity: Low in pass 1, P0 nowcollects Low.severity high means …) now collects a phantom finding at rank 3.Severity: fixablecollects rank 0, not CRITICAL (main dropped the line).#3717 trial merge (head 5d2ba4f): the only conflict is #3717 vs main's #3845
ledgerBlockersblock, not these hunks; resolved by keeping both. #3717's verify tests use JSON findings and a no-fieldBlocks: yesline, so they keep their meaning. Under #3717, verify re-derives labels, so a ledger collected before this change with a field-plus-body-token finding would report merged_line_differs.Review-loop exception: accepted by GStack Minder (GSTA-36) 2026-11-02 via GSTA-36: GSTA-36 2026-11-02: one scoped fix-check on c950a49..f3810d4 (rule 23), cap 1; GSTA-36 2026-11-02: one single-reader scoped check on the post-fix-check repair hunk, cap 3 (fix-check at c9e3cd9, 1 finding)