Repository navigation
docs(credits): record carried work from #3284 (Gemini 3.8 Flash Antigravity support) landed via #3286 #3431
Description
Documentation problem type
Missing documentation
Documentation location
CREDITS.md
What is wrong or missing?
CREDITS.md ("Carried work") is missing an entry for the Google Antigravity Gemini 3.8 Flash tiered wire model collapse and effort ladder design first submitted in PR #3284 and landed on dev via PR #3286 (3d3c4fe26).
In accordance with CREDITS.md ("If you find a landing that belongs on this page, open an issue") and maintainer @Ingwannu's instruction on #3284 ("Please open that focused credit issue referencing #3284, #3286, and the maintainer statement/evidence that identifies what was carried"), this issue records the factual landing evidence for @lidge-jun to make the attribution decision.
What should the documentation explain instead?
CREDITS.md under ## Carried work should record @mdwsk88 as the author of the original implementation in #3284 that was carried/superseded into #3286 (3d3c4fe26).
Suggested wording or example
In the ## Carried work table in CREDITS.md:
| Pull request | Author | Landed as | What landed |
|---|---|---|---|
| #3284 | @mdwsk88 | 3d3c4fe26 |
"Core implementation is already on dev via #3286 (3d3c4fe26), including the suffix wire ladder, picker collapse, Google adapter coverage" — Gemini 3.8 Flash tiered wire models and effort routing on Antigravity |
Additional context or attachments
Factual timeline and maintainer statements:
- 2026-09-02 16:41 UTC: @mdwsk88 opened PR feat(antigravity): support Gemini 3.8 Flash tiered wire models and effort ladder #3284 (
feat(antigravity): support Gemini 3.8 Flash tiered wire models and effort ladder, head4ab777ec6), passing all deterministic hygiene and gate checks. - 2026-09-02 16:44 UTC: Maintainer @lidge-jun reviewed feat(antigravity): support Gemini 3.8 Flash tiered wire models and effort ladder #3284 (priority 64/80, recommending merge): "이 PR는 Google Antigravity(CCA) picker에 gemini-3.8-flash를 추가하고, effort low/medium/high를 wire 접미사 gemini-3.8-flash-low/medium/high로 보내는 사다리와 기본값 medium을 넣는다... 모델 슬라이스는 이 PR(feat(antigravity): support Gemini 3.8 Flash tiered wire models and effort ladder #3284)이 더 깨끗하다."
- 2026-09-02 17:20 UTC: Maintainer @lidge-jun opened PR feat(models): roll out Gemini 3.8 Flash across Antigravity, Google, and the sidecar #3286 (
3d3c4fe26) carrying this implementation alongside pricing/metadata extensions, and merged it todevat 17:48 UTC. - 2026-09-03 09:13 UTC: Maintainer @Ingwannu closed feat(antigravity): support Gemini 3.8 Flash tiered wire models and effort ladder #3284 as superseded: "Core implementation is already on dev via feat(models): roll out Gemini 3.8 Flash across Antigravity, Google, and the sidecar #3286 (3d3c4fe), including the suffix wire ladder, picker collapse, Google adapter coverage... Please open that focused credit issue referencing feat(antigravity): support Gemini 3.8 Flash tiered wire models and effort ladder #3284, feat(models): roll out Gemini 3.8 Flash across Antigravity, Google, and the sidecar #3286, and the maintainer statement/evidence that identifies what was carried."
Checks
- I searched existing documentation issues.
- No secrets or personal information are included.
Activity
lidge-jun commentedon Sep 4, 2026
리뷰 · 우선순위 38 / 80
설명
이 이슈는 코드 버그가 아니라
CREDITS.md의 Carried work 표에 #3284(@mdwsk88) → #3286(3d3c4fe26) 귀속을 남겨 달라는 문서 요청입니다. 본문이CREDITS.md안내(“랜딩이 빠졌으면 이슈로 증거를 열어라”)와 @Ingwannu가 #3284를 superseded로 닫으며 남긴 지시(“focused credit issue… Jun이 사실 귀속을 판단”)를 그대로 따릅니다. 타임라인도 구체적입니다. #3284 오픈·그록봇 리뷰(우선순위 64)·메인테이너 #3286 랜딩·#3284 종료 순서가 적혀 있고, 닫힘 코멘트의 “Core implementation is already on dev via #3286” 문장까지 인용합니다.현재
dev의CREDITS.md## Carried work표를 보면 #3284/#3286/@mdwsk88 행은 아직 없습니다. 랜딩 커밋3d3c4fe26(feat(models): roll out Gemini 3.8 Flash across Antigravity, Google, and the sidecar (#3286))은 HEAD 역사에 있고, Antigravity suffix effort ladder 등이 그 유닛에 포함됩니다. 파일 서문은 “닫는 코멘트나 랜딩 커밋에 나온 메인테이너 말만 인용하고, diff 유사성만으로 추론하지 말라”고 못 박습니다. 그래서 이 이슈의 역할은 표 초안을 제안하는 것이고, 넣있을지 말지는 메인테이너 사실 판단입니다. #3284 본인이 “같은 디자인이 곧 랜딩됐다”고 크레딧을 요청한 기록과, Ingwannu가 “이 PR을 코드로 다시 받지 말고 크레딧 이슈로 분리하라”고 한 기록이 둘 다 있습니다.우선순위 38은 제품 동작·보안과 무관한 귀속 문서이고, 이미 코드는 #3286으로
dev에 있다는 뜻입니다. 빨리 처리할 가치는 있지만 버그 큐를 밀어내지는 않습니다. types/config 분할과도 무관합니다.경로 CREDITS.md
## Carried work- HEAD에 #3284 행이 없습니다. 이 이슈가 채우려는 칸입니다.경로 커밋
3d3c4fe26/ PR #3286 - 랜딩 사실은 확인됩니다. Landed as 칸 후보가 맞습니다.경로 PR #3284 (closed superseded) - 코드 재머지 대상이 아닙니다. 이 이슈만으로 귀속을 결정하면 됩니다.
경로 Report and diagnosis 표 - 여긴 “보고만 살리고 코드는 다른 길”용입니다. #3284는 구현 carry 주장에 가깝습니다.
메인테이너의 판단이 필요한 지점
너의 추천
코드 PR은 받지 마세요. Jun이 #3284↔#3286 증거를 한 번 확인한 뒤, 맞으면 제안 표대로
CREDITS.md한 줄 PR만 머지하고 이 이슈를 닫으세요. 근거가 부족하다고 보면 그 이유를 남기고 wontfix/close 하면 됩니다.이 댓글은 grok-bot이 작성했습니다
feat: close the three verified triage gaps (fable-5-1 metadata, #3284 …
lidge-jun commentedon Sep 4, 2026
Recorded on
devas3a9c4d297(PR #3478).CREDITS.mdnow carries:| #3284 | @mdwsk88 |
3d3c4fe26| "Core implementation is already ondevvia #3286 (3d3c4fe26), including the suffix wire ladder, picker collapse, Google adapter coverage" |The quote is verbatim from @Ingwannu's closing comment on #3284, which is what the file requires — it says entries cite the maintainer's own words and are never inferred from diff similarity.
Every factual claim in this issue was verified independently rather than accepted from the report:
3d3c4fe26is an ancestor oforigin/devand is the #3286 merge, #3284 is closed and not merged, andsrc/providers/antigravity-models.tsondevnamesgemini-3.8-flash16 times — so the suffix ladder that PR argued for is what shipped.Thanks for opening it with the timeline attached; that is exactly the evidence the file asks for, and it made the attribution decision straightforward rather than a judgment call.
docs(devlog): record the triage gap closure and the review round
docs(devlog): record the triage gap closure and the review round (#3482)
feat: close the three verified triage gaps (fable-5-1 metadata, lidge…
docs(devlog): record the triage gap closure and the review round (lid…