Skip to content

feat(quotes): add Mooncake supporter quote / 添加 Mooncake 支持者评价 - #737

Merged
Oseltamivir merged 2 commits into
masterfrom
agent/add-mooncake-supporter-quote
Aug 14, 2026
Merged

feat(quotes): add Mooncake supporter quote / 添加 Mooncake 支持者评价#737
Oseltamivir merged 2 commits into
masterfrom
agent/add-mooncake-supporter-quote

Conversation

@Oseltamivir

@Oseltamivir Oseltamivir commented Aug 14, 2026

Copy link
Copy Markdown
Contributor

Note

Low Risk
Content and asset updates on marketing quotes UI with accompanying tests; no auth, data, or core inference logic changes.

Overview
Adds a Mooncake supporter quote to the quotes page and landing carousel, with English and Simplified Chinese copy attributed to the Mooncake Team and linking to kvcache.ai.

Introduces kvcache-ai.svg for the carousel/logo treatment and extends unit and Cypress coverage so Mooncake appears in CAROUSEL_ORGS and the #quote-mooncake block renders the expected text, link, logo, and zh translation.

Reviewed by Cursor Bugbot for commit 416c8f8. Bugbot is set up for automated code reviews on this repo. Configure here.

Add the Mooncake Team endorsement with Simplified Chinese copy, the official KVCache.AI logo, and focused component coverage.

中文:新增 Mooncake 团队支持者评价及简体中文文案,使用官方 KVCache.AI 标志,并补充针对性的组件测试。
@vercel

vercel Bot commented Aug 14, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
inferencemax-app Ready Ready Preview Aug 14, 2026 3:26pm

Request Review

Add Mooncake to the landing-page carousel membership and cover it with a regression test.

中文:将 Mooncake 加入首页支持者轮播,并补充回归测试以确保其持续展示。
@Oseltamivir
Oseltamivir marked this pull request as ready for review August 14, 2026 15:28
@Oseltamivir
Oseltamivir requested a review from adibarra as a code owner August 14, 2026 15:28
@Oseltamivir
Oseltamivir merged commit 844df76 into master Aug 14, 2026
24 of 31 checks passed
@Oseltamivir
Oseltamivir deleted the agent/add-mooncake-supporter-quote branch August 14, 2026 15:28
Oseltamivir added a commit that referenced this pull request Aug 15, 2026
Pareto frontier from its selected axes without E2E normalized-interactivity
gating" — and shipped with "full checks intentionally skipped for this focused
change". Master has been red since: 3 unit tests and 3 e2e suites still assert
the gating, and the api-doc digest gate was left tripped. This updates the tests
to the behaviour that was chosen. No src behaviour changes.

Unit:

- overview-data: an agentic point dominated on total tokens and slower on E2E is
  no longer dropped, so the tier-50 read lands on it (8100) instead of coming
  back null.
- useThroughputData: buildGpuGroups keeps an e2e-dominated point, because
  eligibility is now decided downstream from the selected axes.
- api-route-catalog: reviewed overview-data.ts for #736 and updated its digest.
  Tier reads change value, but the Overview BFF parameters and the
  OverviewPageData shape are byte-identical — verified by diffing the exported
  surface against 08e1f7d — so no published contract moved.

E2E:

- The "Advanced" x-axis popover is gone and all four modes are flat tabs, so the
  four specs that opened it now click the tab. Agentic also defaults to
  Interactivity rather than E2E Normalized Interactivity.
- The derived-metrics stub moves to beforeEach. Cypress clears intercepts
  between tests, and these specs relied on the old default fetching during
  `before` while the stub was alive; React Query then held the result for the
  spec. Nothing fetches on load now, so a later switch into that mode issued a
  live request and rendered no chart.
- The URL-restored-mode regression test now restores TTFT. It asked for
  Interactivity, which is the new default, so it could no longer fail.
- overlay-optimal-only asserted that overlay points are hidden because
  trace-less rows cannot join the canonical frontier. They now stand or fall on
  the selected axes like any other point, and all five fixture configs are
  non-dominated there — throughput falls monotonically as interactivity rises —
  so the suite would have asserted five visible either way. It gets a sixth,
  deliberately dominated overlay config, so Optimal Only has something to drop;
  making that point non-dominated fails the test.

Verified: 3363 unit tests, 598/602 e2e locally. Two local failures are not from
this change — zh-pages passes in isolation (page-load timeout under parallel
load), and inference-replay fails on a wall-clock animation assertion, imports
nothing touched here, and passed CI on #737.

中文:#736 有意移除了端到端归一化交互性的前沿门控,并"刻意跳过完整检查",导致 master
持续为红:3 项单测与 3 个 e2e 套件仍在断言已被移除的门控,API 文档摘要闸门也未更新。
本提交只把测试对齐到既定行为,不改动任何源码行为。单测:概览档位读数不再因总 token
被支配而丢弃该点(改为 8100);buildGpuGroups 保留被 e2e 支配的点;复核 overview-data.ts
后更新摘要(对外参数与响应结构逐字节一致)。E2E:Advanced 弹出菜单已移除,四个模式改为
平铺标签页;派生指标 stub 移至 beforeEach(Cypress 会在测试间清除拦截);URL 恢复模式
的回归测试改用 TTFT(原用 Interactivity 已与新默认值相同而无法失败);overlay-optimal-only
新增一个刻意被支配的配置,否则该断言无论过滤器如何都会通过。
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant