v1.89.0.0 feat: add shared-code extraction audit (#2925)

* feat: bind shared-code review advice to source and branch

* feat: add shared-code extraction audit and scoped review checks

* test: recognize complete source reads and explicit coverage legends

* chore: bump version and changelog (v1.88.0.0)

Co-Authored-By: OpenAI Codex <noreply@openai.com>

* test: capture native review questions and retain public evidence

Capture the actual first public native question with strict ownership and display matching. Preserve terminal failures and raw evidence, and retain SDK completion checks.

* test: recognize verified review evidence and complete fixtures

Recognize complete source and diagram evidence, concrete design and developer-experience decisions, and the complete planted scenario contracts. Preserve negative controls and grading thresholds.

* fix: preserve decision brief structure in native questions

Keep the required pros-and-cons heading and final Net field in native question text. Regenerate host outputs and document the release and evaluation repairs.

Co-Authored-By: OpenAI Codex <noreply@openai.com>

* docs: update project documentation for v1.88.0.0

Co-Authored-By: OpenAI Codex <noreply@openai.com>

* fix: correct eval retry accounting and ship workflow gates

* fix: capture native eval evidence and stabilize CI fixtures

* fix: keep shared-code eval skips read-only

Choose explicit no-change answers instead of mixed fix/preservation options.
Reuse the bounded revalidation prompt for path fixtures so required review
metadata is available without repeated discovery. Preserve source checks,
retry limits, and failed native terminal outcomes.

Add captured-question and callback regressions, plus evaluation selection
coverage for the affected fixtures.

---------

Co-authored-by: OpenAI Codex <noreply@openai.com>
This commit is contained in:
Garry Tan
2026-09-24 01:53:58 -04:00
committed by GitHub
co-authored by OpenAI Codex
parent b9706f3635
commit 06ed920a97
177 changed files with 13244 additions and 2477 deletions
+43
View File
@@ -0,0 +1,43 @@
{
"provenance": {
"source": "September 21 bounded plan-design-finding-count confirmation",
"runId": "shared-libs-design-count-sep21",
"observedReviewCount": 0,
"observedStep0Count": 6,
"sourceCallSha256": "e56b9c057b1a69193d2a932eafac5ca81c8dbe8ed21009d147adc59c881b2052"
},
"calls": [
{
"sessionId": "1a7d5895-da78-46d8-9fa9-e8b681038344",
"toolUseId": "toolu_01Fes64xV7Kq6QiS9HH49Drc",
"questions": [
{
"question": "D1 — Issue 1: Save is visually identical to Reset, Cancel, and Export. How should the primary action be distinguished?\nProject/branch/task: gstack-plan-count-PXjOyl on main, design review of PLAN.md (Settings Page UI redesign).\nELI10: Four buttons in a row all look the same, so the person has to read every label to find the one they came for. Making Save the only filled button turns a reading task into a glance. The stakes: mis-taps on Reset or Cancel open a discard dialog, and the page reads as unfinished.\nStakes if we pick wrong: users hesitate at the action row or hit a secondary action, and the redesign ships looking like the old form.\nRecommendation: 1A because DESIGN.md already names the token (filled #1d4ed8 with white text, ghosts for the rest) and the existing Button component supports it. Principle: hierarchy as service. \nCompleteness: A=10/10, B=6/10, C=3/10\nNet: 1A applies the approved system; 1B and 1C invent something DESIGN.md does not sanction.",
"header": "Save primary",
"multiSelect": false,
"options": [
{
"label": "1A: Filled Save, ghost others (recommended)",
"description": "✅ Matches DESIGN.md exactly: Save filled #1d4ed8 with white text, Reset/Cancel/Export neutral ghost buttons. ✅ Uses existing Button variants; no new styles, 44px geometry and focus ring unchanged. ❌ Ghost buttons rely on the existing Button component's neutral text color meeting AA; verify in QA. (human: ~1h / CC: ~5min)"
},
{
"label": "1B: Filled Save plus visual divider before Export",
"description": "✅ Also separates the non-destructive Export from the two discard actions. ✅ Keeps DOM and tab order intact. ❌ Adds a divider element DESIGN.md does not define, so it needs its own token and responsive rule at 640px. ❌ More chrome on an already busy header row. (human: ~2h / CC: ~10min)"
},
{
"label": "1C: Keep four equal buttons, bold the Save label only",
"description": "✅ Smallest change to the existing markup and stylesheet. ✅ No color work, so no contrast verification needed. ❌ Weight alone is a weak signal at a glance and violates DESIGN.md, which names Save the only filled action. ❌ Leaves the primary action undiscoverable for scanning users. (human: ~20min / CC: ~2min)"
}
]
}
],
"answered": true,
"failed": false,
"answers": {
"D1 — Issue 1: Save is visually identical to Reset, Cancel, and Export. How should the primary action be distinguished?\nProject/branch/task: gstack-plan-count-PXjOyl on main, design review of PLAN.md (Settings Page UI redesign).\nELI10: Four buttons in a row all look the same, so the person has to read every label to find the one they came for. Making Save the only filled button turns a reading task into a glance. The stakes: mis-taps on Reset or Cancel open a discard dialog, and the page reads as unfinished.\nStakes if we pick wrong: users hesitate at the action row or hit a secondary action, and the redesign ships looking like the old form.\nRecommendation: 1A because DESIGN.md already names the token (filled #1d4ed8 with white text, ghosts for the rest) and the existing Button component supports it. Principle: hierarchy as service. \nCompleteness: A=10/10, B=6/10, C=3/10\nNet: 1A applies the approved system; 1B and 1C invent something DESIGN.md does not sanction.": "1A: Filled Save, ghost others (recommended)"
},
"unansweredQuestionIndices": [],
"answeredAt": "2026-09-21T17:11:29.350Z"
}
]
}