From 0a2c0a92acca59b772b06133b1594fbc51e83aed Mon Sep 17 00:00:00 2001
From: Chinesezjc <jczhai@deepseek.com>
Date: Mon, 10 Aug 2026 22:59:14 +0800
Subject: [PATCH] ci: address review feedback for Windows CI unification

- Rewrite ci-workflow.spec.ts to assert new single-job native Windows topology
- Remove check:windows-wine from root AGENTS.md command list
- Archive 2026-08-08-native-windows-pull-request-ci (dual-lane topology
  superseded) with Archived: metadata and inbound-link updates
- Update cross-references in serial-cross-platform-ci-reference,
  portable-required-pull-request-ci, and pnpm-action-setup notes
- Add Windows pool coverage to the failover runbook (EN + zh)
- Remove dead wine-windows-gates.sh
- Remove unnecessary Playwright Chromium install from serial-windows
- Fix .github/AGENTS.md windows-* label wording
- Fix PR description before-state accuracy
---
 ...8-native-windows-pull-request-ci.i18n.yaml |   6 +
 ...26-08-08-native-windows-pull-request-ci.md |   1 +
 ...08-08-native-windows-pull-request-ci.zh.md |   1 +
 ...rial-cross-platform-ci-reference.i18n.yaml |   4 +-
 ...7-21-serial-cross-platform-ci-reference.md |   4 +-
 ...1-serial-cross-platform-ci-reference.zh.md |   4 +-
 ...ortable-required-pull-request-ci.i18n.yaml |   4 +-
 ...07-23-portable-required-pull-request-ci.md |   6 +-
 ...23-portable-required-pull-request-ci.zh.md |   6 +-
 .../2026-07-26-ci-failover-runbook.i18n.yaml  |   4 +-
 .../process/2026-07-26-ci-failover-runbook.md |   8 +-
 .../2026-07-26-ci-failover-runbook.zh.md      |   8 +-
 ...n-setup-for-symmetric-ci-caching.i18n.yaml |   4 +-
 ...m-action-setup-for-symmetric-ci-caching.md |   4 +-
 ...ction-setup-for-symmetric-ci-caching.zh.md |   4 +-
 ...8-native-windows-pull-request-ci.i18n.yaml |   6 -
 .github/AGENTS.md                             |   2 +-
 .github/workflows/ci.yml                      |   7 -
 AGENTS.md                                     |   1 -
 scripts/ci-workflow.spec.ts                   |  56 ++--
 scripts/wine-windows-gates.sh                 | 282 ------------------
 21 files changed, 78 insertions(+), 344 deletions(-)
 create mode 100644 .agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml
 rename .agents/notes/{implemented => archived}/process/2026-08-08-native-windows-pull-request-ci.md (99%)
 rename .agents/notes/{implemented => archived}/process/2026-08-08-native-windows-pull-request-ci.zh.md (99%)
 delete mode 100644 .agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml
 delete mode 100755 scripts/wine-windows-gates.sh

diff --git a/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml b/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml
new file mode 100644
index 0000000000..0a565b17d1
--- /dev/null
+++ b/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write .agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.md
+2026-08-08-native-windows-pull-request-ci.md: 3f69f335f6d873124fbbc1fd96d06ff2b3c8dc43
+2026-08-08-native-windows-pull-request-ci.zh.md: 136a7e8791da141cc12ba77d148f47b28db4790b
diff --git a/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.md b/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.md
similarity index 99%
rename from .agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.md
rename to .agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.md
index 6a62fddb79..3f69f335f6 100644
--- a/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.md
+++ b/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.md
@@ -1,6 +1,7 @@
 # Agent Note: Dual Wine and native Windows pull-request CI
 
 Status: implemented
+Archived: 2026-08-10
 
 English | [中文](2026-08-08-native-windows-pull-request-ci.zh.md)
 
diff --git a/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.zh.md b/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.zh.md
similarity index 99%
rename from .agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.zh.md
rename to .agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.zh.md
index 990b8ed143..136a7e8791 100644
--- a/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.zh.md
+++ b/.agents/notes/archived/process/2026-08-08-native-windows-pull-request-ci.zh.md
@@ -1,6 +1,7 @@
 # Agent Note: Wine 与原生 Windows 双通道拉取请求 CI
 
 Status: implemented
+Archived: 2026-08-10
 
 [English](2026-08-08-native-windows-pull-request-ci.md) | 中文
 
diff --git a/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.i18n.yaml b/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.i18n.yaml
index 33bffe92ff..a65b22ba3f 100644
--- a/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.i18n.yaml
+++ b/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.md
-2026-07-21-serial-cross-platform-ci-reference.md: 07dc430e6fed3fe75a006ca03523bd6e4fc969d0
-2026-07-21-serial-cross-platform-ci-reference.zh.md: 8bbb60cdead2957069de22ecaddf01c6cd9fb305
+2026-07-21-serial-cross-platform-ci-reference.md: b7b90f84095fe317f92c051f66da0bf8ee080b19
+2026-07-21-serial-cross-platform-ci-reference.zh.md: d62d7d754eb0343d8b1dd86adae7f501c04b177b
diff --git a/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.md b/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.md
index 07dc430e6f..b7b90f8409 100644
--- a/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.md
+++ b/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.md
@@ -16,7 +16,7 @@ Real-kernel sandbox proofs require specific hosted operating systems and archite
 
 ## Decision
 
-[CI](../../../../.github/workflows/ci.yml) gives pull-request and master-push events complementary responsibilities. Pull requests run consolidated Linux and Wine-hosted Windows jobs plus the Node compatibility and Python contracts on standard GitHub-hosted capacity; an independent native Windows job reports the complete Windows inventory without participating in the required aggregate. On a push to `master`, the active reference is `serial / linux (self-hosted standby)` on the in-house `vm-backup` pool — the hot-standby drill that continuously re-proves the failover target described in the [failover runbook](2026-07-26-ci-failover-runbook.md). The standard-hosted `serial / linux`, `serial / macos`, and `serial / windows` definitions remain disabled under `TODO(hosted-serial-ci)` until their portable capacity can be restored. The separate job definitions intentionally keep their short checkout, runtime setup, and immutable install sequences visible instead of hiding operating systems behind a matrix or reusable workflow. `workflow_dispatch` is reserved for runner benchmarks.
+[CI](../../../../.github/workflows/ci.yml) gives pull-request and master-push events complementary responsibilities. Pull requests run consolidated Linux and native Windows jobs plus the Node compatibility and Python contracts. On a push to `master`, the active references are `serial / linux (self-hosted standby)` and `serial / windows (self-hosted standby)` on the in-house self-hosted pools — the hot-standby drills that continuously re-prove the failover targets described in the [failover runbook](2026-07-26-ci-failover-runbook.md). The standard-hosted `serial / linux`, `serial / macos`, and `serial / windows` definitions remain disabled under `TODO(hosted-serial-ci)` until their portable capacity can be restored. The separate job definitions intentionally keep their short checkout, runtime setup, and immutable install sequences visible instead of hiding operating systems behind a matrix or reusable workflow. `workflow_dispatch` is reserved for runner benchmarks.
 
 Each reference job runs `pnpm run check:ci` without any shard selector. `DSH_GATE_CONCURRENCY=1` makes the top-level aggregate execute one ready gate at a time; coverage, snapshot replay, built-bin smoke, and publication validation also receive worker counts of one. The reference jobs may run beside one another, but each host's repository gates are serial and complete. Linux installs bubblewrap before replaying snapshots, and Windows enables Developer Mode before installing the symlinked workspace.
 
@@ -28,7 +28,7 @@ The standalone [Sandbox](../../../../.github/workflows/sandbox.yml) workflow bel
 
 Master reference jobs are diagnostic and do not participate in the pull request's required `all checks passed` result. The CI and Sandbox workflows keep their cross-platform references on master pushes. Performance is evaluated from completed hosted-job timestamps and reported as a measurement; it is not encoded as a `timeout-minutes` value.
 
-The portable reference uses GitHub's standard `ubuntu-latest`, `macos-latest`, and `windows-2025` labels. The required pull-request Windows job runs under Wine on `ubuntu-latest`, while the independent pull-request native job uses standard `windows-2025` under the [dual Windows decision](2026-08-08-native-windows-pull-request-ci.md); when enabled, `serial / windows` remains a second complete, unsharded native-kernel oracle. Required pull-request jobs use portable standard capacity under the [required-CI decision](2026-07-23-portable-required-pull-request-ci.md). Higher-core hosted runners remain manual benchmarks because a correctness path must remain runnable without repository-external runner configuration.
+The portable reference uses GitHub's standard `ubuntu-latest`, `macos-latest`, and `windows-2025` labels. The required pull-request Windows job runs native on the enterprise larger runner under normal operation, failing over to the self-hosted pool under `DSH_CI_FAILOVER=selfhosted`; the archived [dual Windows decision](../archived/process/2026-08-08-native-windows-pull-request-ci.md) records the prior Wine/native topology. Required pull-request jobs use portable standard capacity under the [required-CI decision](2026-07-23-portable-required-pull-request-ci.md). Higher-core hosted runners remain manual benchmarks because a correctness path must remain runnable without repository-external runner configuration.
 
 ## Alternatives considered
 
diff --git a/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.zh.md b/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.zh.md
index 8bbb60cdea..d62d7d754e 100644
--- a/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.zh.md
+++ b/.agents/notes/implemented/process/2026-07-21-serial-cross-platform-ci-reference.zh.md
@@ -16,7 +16,7 @@ Status: implemented
 
 ## 决策
 
-[CI](../../../../.github/workflows/ci.yml) 为拉取请求事件与 master 推送事件赋予互补的职责。拉取请求在 GitHub 标准托管容量上运行合并后的 Linux 和由 Wine 承载的 Windows 作业，以及 Node 兼容性与 Python 约定；一个独立的原生 Windows 作业会报告完整的 Windows 清单，但不参与必需聚合流程。向 `master` 推送时，当前启用的参考作业是公司自有 `vm-backup` 池上的 `serial / linux (self-hosted standby)`——该热备演练持续验证[故障切换手册](2026-07-26-ci-failover-runbook.md)所描述的切换目标。标准托管的 `serial / linux`、`serial / macos` 和 `serial / windows` 定义仍处于禁用状态，并由 `TODO(hosted-serial-ci)` 标记，直到其可移植容量恢复。各自独立的作业定义有意显式保留简短的代码检出、运行时设置和依赖锁定的安装步骤，而不是用矩阵或可复用工作流隐藏操作系统差异。`workflow_dispatch` 仅用于运行器基准测试。
+[CI](../../../../.github/workflows/ci.yml) 为拉取请求事件与 master 推送事件赋予互补的职责。拉取请求运行合并后的 Linux 与原生 Windows 作业，以及 Node 兼容性与 Python 约定。向 `master` 推送时，当前启用的参考作业是公司自有自托管池上的 `serial / linux (self-hosted standby)` 与 `serial / windows (self-hosted standby)`——这些热备演练持续验证[故障切换手册](2026-07-26-ci-failover-runbook.md)所描述的切换目标。标准托管的 `serial / linux`、`serial / macos` 和 `serial / windows` 定义仍处于禁用状态，并由 `TODO(hosted-serial-ci)` 标记，直到其可移植容量恢复。各自独立的作业定义有意显式保留简短的代码检出、运行时设置和依赖锁定的安装步骤，而不是用矩阵或可复用工作流隐藏操作系统差异。`workflow_dispatch` 仅用于运行器基准测试。
 
 每个参考作业均在不设置任何分片选择器的情况下运行 `pnpm run check:ci`。`DSH_GATE_CONCURRENCY=1` 使顶层聚合每次只执行一个已经就绪的门禁；覆盖率、快照回放、built-bin 冒烟测试和发布验证的 worker 数量也设为 1。各参考作业可以彼此并行，但每台主机上的仓库门禁都串行运行且完整执行。Linux 在回放快照前安装 bubblewrap，Windows 则在安装采用符号链接的工作区前启用开发人员模式。
 
@@ -28,7 +28,7 @@ macOS 参考流程使用 fork 进程运行常规 Vitest 项目。macOS arm64 上
 
 master 分支的参考作业仅用于诊断，不参与拉取请求所要求的 `all checks passed` 结果。CI 与 Sandbox 工作流把跨平台参考流程保留在 master 推送上。系统根据已完成托管作业的时间戳评估性能，并将其报告为测量结果，而不是写成 `timeout-minutes` 值。
 
-可移植的参考流程使用 GitHub 标准的 `ubuntu-latest`、`macos-latest` 和 `windows-2025` 标签。拉取请求必需的 Windows 作业在 `ubuntu-latest` 上通过 Wine 运行，而独立的拉取请求原生作业依据[双 Windows 决策](2026-08-08-native-windows-pull-request-ci.md)使用标准 `windows-2025`；`serial / windows` 启用时，仍作为第二个完整且未分片的原生内核标尺。依据[必需 CI 决策](2026-07-23-portable-required-pull-request-ci.md)，拉取请求必需作业使用可移植的标准容量。更高核心数的托管运行器仍仅用于手动基准测试，因为正确性路径必须无需仓库外部的运行器配置即可运行。
+可移植的参考流程使用 GitHub 标准的 `ubuntu-latest`、`macos-latest` 和 `windows-2025` 标签。拉取请求必需的 Windows 作业在正常运行下使用企业级大型运行器，在 `DSH_CI_FAILOVER=selfhosted` 时切换至自托管池；已归档的[双 Windows 决策](../archived/process/2026-08-08-native-windows-pull-request-ci.md)记录此前的 Wine/原生拓扑。依据[必需 CI 决策](2026-07-23-portable-required-pull-request-ci.md)，拉取请求必需作业使用可移植的标准容量。更高核心数的托管运行器仍仅用于手动基准测试，因为正确性路径必须无需仓库外部的运行器配置即可运行。
 
 ## 曾考虑的替代方案
 
diff --git a/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.i18n.yaml b/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.i18n.yaml
index e67565f381..4738f87fda 100644
--- a/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.i18n.yaml
+++ b/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.md
-2026-07-23-portable-required-pull-request-ci.md: b7c637408afa32611c4e70073ddec9aaabfba7f5
-2026-07-23-portable-required-pull-request-ci.zh.md: ee88412d982da5de613d8dfdbdcdd9977c899d2d
+2026-07-23-portable-required-pull-request-ci.md: 6f7a0e0e6354e9bea632d1b00881a44acd921f9e
+2026-07-23-portable-required-pull-request-ci.zh.md: 0ec92f11efa55ed155ada1ba858cd1878398f1ec
diff --git a/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.md b/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.md
index b7c637408a..6f7a0e0e63 100644
--- a/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.md
+++ b/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.md
@@ -12,9 +12,9 @@ Billing health, a runner definition's `Ready` state, and a large autoscaling cei
 
 ## Decision
 
-[CI](../../../../.github/workflows/ci.yml) runs the required primary Node 24 jobs, plus the stable `all checks passed` aggregate, on repo-restricted enterprise 32-core pools. The aggregate performs no checkout or repository gate, but sharing the enterprise pool prevents the required verdict from introducing a separate standard-hosted billing dependency after its substantive jobs have already succeeded. The required Windows job runs Windows Node under Wine on standard `ubuntu-latest` for the blocking surfaces; an independent native `windows-2025` job starts automatically but does not participate in the aggregate ([dual Windows decision](2026-08-08-native-windows-pull-request-ci.md)). Standard `ubuntu-latest` jobs retain Node 22.19, Node 26, and Python SDK compatibility, and the serial references remain the complete unsharded cross-platform definitions. Those standard-hosted jobs keep the portable execution boundary observable without duplicating the primary inventory on every pull request.
+[CI](../../../../.github/workflows/ci.yml) runs the required primary Node 24 jobs, plus the stable `all checks passed` aggregate, on repo-restricted enterprise 32-core pools. The aggregate performs no checkout or repository gate, but sharing the enterprise pool prevents the required verdict from introducing a separate standard-hosted billing dependency after its substantive jobs have already succeeded. The required Windows job runs native on the enterprise larger runner under normal operation, failing over to the self-hosted pool under `DSH_CI_FAILOVER=selfhosted` ([unified Windows CI note](2026-08-10-unified-windows-selfhosted-ci.md)). Standard `ubuntu-latest` jobs retain Node 22.19, Node 26, and Python SDK compatibility, and the serial cross-platform standbys continuously validate the failover targets. Those standard-hosted jobs keep the portable execution boundary observable without duplicating the primary inventory on every pull request.
 
-The two Linux primary jobs, Node compatibility, Python SDK, and `windows node 24 / wine blocking` remain dependencies of `all checks passed`; `windows node 24 / native complete` is deliberately absent. Branch protection continues to require `e2e` and `all checks passed`. There is no automatic fallback when a remaining enterprise Linux label cannot allocate: the standard jobs continue to report their own contracts, but they cannot manufacture the missing required result.
+The two Linux primary jobs, Node compatibility, Python SDK, and `windows node 24 / native complete` remain dependencies of `all checks passed`. The master-only `serial / windows (self-hosted standby)` is deliberately absent from the aggregate, mirroring the Linux standby pattern.
 
 The [larger-runner decision](2026-07-22-evidence-based-larger-hosted-runners.md) owns the current primary topology and its measurements. The [serial cross-platform reference](2026-07-21-serial-cross-platform-ci-reference.md) remains the independent standard-hosted completeness check, and the manual larger-runner suites retain size comparisons without expanding the ordinary required matrix.
 
@@ -30,6 +30,6 @@ The [larger-runner decision](2026-07-22-evidence-based-larger-hosted-runners.md)
 
 ## Consequences
 
-Ordinary pull requests spend enterprise capacity on the Linux critical path while the Wine job keeps the required Windows verdict on standard Linux allocation. The independent native job uses standard Windows allocation without delaying or changing the aggregate. A live exact-head run distinguishes the commands branch protection consumes from the separate diagnostic contract; queue delay is reported separately from each job's `startedAt` to `completedAt` execution interval.
+Ordinary pull requests spend enterprise capacity on the Linux critical path while the native Windows job uses the enterprise larger runner on the normal path and the self-hosted pool under failover. A live exact-head run distinguishes the commands branch protection consumes from the separate diagnostic contract; queue delay is reported separately from each job's `startedAt` to `completedAt` execution interval.
 
 Standard compatibility, required Wine, and diagnostic native Windows jobs remain useful when enterprise allocation is degraded, but they do not make a blocked required Linux job or aggregate green. Recovering Linux availability may require restoring the complete standard-hosted topology; changing a pool definition's status alone is insufficient evidence that it can receive work.
diff --git a/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.zh.md b/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.zh.md
index ee88412d98..0ec92f11ef 100644
--- a/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.zh.md
+++ b/.agents/notes/implemented/process/2026-07-23-portable-required-pull-request-ci.zh.md
@@ -12,9 +12,9 @@ Status: implemented
 
 ## 决策
 
-[CI](../../../../.github/workflows/ci.yml) 在仅限本仓库使用的企业级 32 核运行器池上运行必需的主 Node 24 作业，以及稳定的 `all checks passed` 聚合流程。该聚合流程不执行代码检出或仓库门禁；但让它与所依赖的实质性作业共用企业级运行器池，可以避免这些作业已经成功后，必需判定结果又引入一项单独的标准托管计费依赖。必需的 Windows 作业在标准 `ubuntu-latest` 上通过 Wine 运行 Windows Node，覆盖阻断性检查范围；一个独立的原生 `windows-2025` 作业会自动启动，但不参与聚合流程（[双 Windows 决策](2026-08-08-native-windows-pull-request-ci.md)）。标准 `ubuntu-latest` 作业保留 Node 22.19、Node 26 和 Python SDK 兼容性，串行参考流程仍是完整且未分片的跨平台定义。这些标准托管作业让可移植执行边界保持可观测，而不必在每个拉取请求中重复主清单。
+[CI](../../../../.github/workflows/ci.yml) 在仅限本仓库使用的企业级 32 核运行器池上运行必需的主 Node 24 作业，以及稳定的 `all checks passed` 聚合流程。该聚合流程不执行代码检出或仓库门禁；但让它与所依赖的实质性作业共用企业级运行器池，可以避免这些作业已经成功后，必需判定结果又引入一项单独的标准托管计费依赖。必需的 Windows 作业在正常运行下使用企业级大型运行器，在 `DSH_CI_FAILOVER=selfhosted` 时切换至自托管池（[统一 Windows CI 笔记](2026-08-10-unified-windows-selfhosted-ci.md)）。标准 `ubuntu-latest` 作业保留 Node 22.19、Node 26 和 Python SDK 兼容性，跨平台串行备援持续验证故障切换目标。这些标准托管作业让可移植执行边界保持可观测，而不必在每个拉取请求中重复主清单。
 
-两项 Linux 主作业、Node 兼容性、Python SDK 和 `windows node 24 / wine blocking` 继续作为 `all checks passed` 的依赖项；`windows node 24 / native complete` 被刻意排除。分支保护继续要求 `e2e` 和 `all checks passed`。剩余的企业级 Linux 运行器标签无法分配运行器时没有自动后备机制：标准作业会继续报告各自的约定，但无法产出缺失的必需结果。
+两项 Linux 主作业、Node 兼容性、Python SDK 和 `windows node 24 / native complete` 继续作为 `all checks passed` 的依赖项；仅 master 触发的 `serial / windows (self-hosted standby)` 被刻意排除，与 Linux 备援模式对称。
 
 当前主拓扑及其测量结果由[大型运行器决策](2026-07-22-evidence-based-larger-hosted-runners.md)记录。[跨平台串行参考流程](2026-07-21-serial-cross-platform-ci-reference.md)继续作为独立的标准托管完整性检查，手动大型运行器套件则保留规格比较，同时不扩大普通必需矩阵。
 
@@ -30,6 +30,6 @@ Status: implemented
 
 ## 后果
 
-普通拉取请求会将企业级运行器容量用于 Linux 关键路径，而 Wine 作业让必需的 Windows 判定继续使用标准 Linux 运行器容量。独立原生作业使用标准 Windows 运行器容量，不会延迟或改变聚合流程。一次实际的分支头精确运行会区分分支保护采用的命令与单独的诊断约定；排队延迟与每个作业从 `startedAt` 到 `completedAt` 的执行区间分开报告。
+普通拉取请求会将企业级运行器容量用于 Linux 关键路径，而原生 Windows 作业在正常运行下使用企业级大型运行器、故障切换时使用自托管池。独立原生作业使用标准 Windows 运行器容量，不会延迟或改变聚合流程。一次实际的分支头精确运行会区分分支保护采用的命令与单独的诊断约定；排队延迟与每个作业从 `startedAt` 到 `completedAt` 的执行区间分开报告。
 
 企业级运行器分配能力下降时，标准兼容性作业、必需的 Wine 作业与诊断性原生 Windows 作业仍能提供有用证据，但无法让受阻的必需 Linux 作业或聚合流程变绿。恢复 Linux 可用性时，可能需要恢复完整的标准托管拓扑；仅改变运行器池定义的状态，不足以证明它可以接收作业。
diff --git a/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.i18n.yaml b/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.i18n.yaml
index f4fe0c74d2..2403d869c6 100644
--- a/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.i18n.yaml
+++ b/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/process/2026-07-26-ci-failover-runbook.md
-2026-07-26-ci-failover-runbook.md: 72261f95ea74b61e3915a1a6419b2c2e616efbd9
-2026-07-26-ci-failover-runbook.zh.md: 1ee679eabc296ab31d71b87945409788539425bd
+2026-07-26-ci-failover-runbook.md: 25bfb8cc7ca6d53c5d28e7c941facdb86e5e147b
+2026-07-26-ci-failover-runbook.zh.md: 22195500f5a28c9e92bc8aac8c58028628ab1d6b
diff --git a/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.md b/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.md
index 72261f95ea..25bfb8cc7c 100644
--- a/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.md
+++ b/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.md
@@ -6,16 +6,20 @@ English | [中文](2026-07-26-ci-failover-runbook.zh.md)
 
 ## Problem
 
-The three required Linux worker jobs in [CI](../../../../.github/workflows/ci.yml) (`node 24 / static`, `node 24 / coverage`, `node 24 / snapshots and artifacts`) run on the hosted enterprise 32-core pools; the required verdict job that aggregates them (`all checks passed`) runs on standard `ubuntu-latest`. When the enterprise pools degrade — jobs queue indefinitely or the enterprise labels vanish — every open pull request becomes unmergeable, and the ordinary recovery of merging a fix is itself deadlocked behind the very required checks that cannot run. **Scope: this switch recovers an enterprise Linux-pool outage.** The verdict's other required dependencies (`node-compat`, `python-sdk`, `windows`) stay on standard hosted runners by design (the portable boundary); in a broader GitHub-hosted capacity failure that also takes out the standard pools, those dependencies still block `all checks passed`, and only the Windows leg has no in-house substitute at all — during the 2026-07-27 outage the standard pools recovered first, which is the ordering this design bets on. An outage therefore needs a switch any responder with repository write access can throw without merging anything.
+The three required Linux worker jobs in [CI](../../../../.github/workflows/ci.yml) (`node 24 / static`, `node 24 / coverage`, `node 24 / snapshots and artifacts`) run on the hosted enterprise 32-core pools; the required verdict job that aggregates them (`all checks passed`) runs on standard `ubuntu-latest`. The required Windows job (`windows node 24 / native complete`) runs on the hosted enterprise larger runner. When the enterprise pools degrade — jobs queue indefinitely or the enterprise labels vanish — every open pull request becomes unmergeable, and the ordinary recovery of merging a fix is itself deadlocked behind the very required checks that cannot run. **Scope: this switch recovers an enterprise Linux-pool outage AND a hosted Windows-pool outage.** The verdict's other required dependencies (`node-compat`, `python-sdk`) stay on standard hosted runners by design (the portable boundary); in a broader GitHub-hosted capacity failure that also takes out the standard pools, those dependencies still block `all checks passed`. An outage therefore needs a switch any responder with repository write access can throw without merging anything.
 
 ## Decision
 
-Each of the three required Linux worker jobs — and the `all checks passed` verdict job, which would otherwise stay queued on the failed pool even after every worker passed — resolves its runner pool through the `DSH_CI_FAILOVER` repository variable. Unset (normal), they run on the hosted enterprise pools. Set to `selfhosted` by any repository writer, all four retarget onto the in-house self-hosted `vm-backup` pool, coverage and snapshot concurrency drop to shared-VM bounds, and the hosted-path pnpm cache restores are skipped. The switch is writer-manageable repository state, not a merge, so it works while every check is red. The in-house pool's readiness is continuously re-proven by the `serial / linux (self-hosted standby)` lane, which runs the complete unsharded aggregate on every master push.
+Each of the three required Linux worker jobs, the required Windows job, and the `all checks passed` verdict job — which would otherwise stay queued on the failed pool even after every worker passed — resolves its runner pool through the `DSH_CI_FAILOVER` repository variable. Unset (normal), they run on the hosted enterprise pools. Set to `selfhosted` by any repository writer, all five retarget onto the in-house self-hosted pools: the Linux jobs and verdict onto the `vm-backup` pool (`[self-hosted, linux, x64, vm-backup]`), coverage and snapshot concurrency drop to shared-VM bounds, and the hosted-path pnpm cache restores are skipped; the Windows job onto the `dsh-win-ci` pool (`[self-hosted, dsh-win-ci, windows]`). The switch is writer-manageable repository state, not a merge, so it works while every check is red. The in-house pools' readiness is continuously re-proven by the `serial / linux (self-hosted standby)` and `serial / windows (self-hosted standby)` lanes, which run the complete unsharded aggregates on every master push.
 
 ### What the in-house pool is
 
 `vm-backup`: one 64-core VM, six always-on systemd-managed runner instances. Its image must preinstall Playwright Chromium's Linux system packages; CI downloads the lockfile-selected browser but never runs `apt` on this persistent shared host. Check the latest `serial / linux (self-hosted standby)` run before switching: its aggregate includes browser replay, so a green standby verifies both ordinary capacity and this browser prerequisite.
 
+#### Windows pool
+
+`dsh-win-ci`: 16 always-on runner instances on the in-house Windows CI server. Labels: `[self-hosted, dsh-win-ci, windows]`. The image must preinstall Node 24, pnpm, Git, and enable Developer Mode for symlink support; Playwright Chromium is installed per-run by the `serial / windows (self-hosted standby)` lane. Check the latest `serial / windows (self-hosted standby)` run before switching: a green standby verifies the pool can execute `check:ci:windows-complete` end-to-end.
+
 ### Switch (any repository writer, ~1 minute, no merge)
 
 1. Repository **Settings → Secrets and variables → Actions → Variables → New repository variable**: name `DSH_CI_FAILOVER`, value `selfhosted`.
diff --git a/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.zh.md b/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.zh.md
index 1ee679eabc..22195500f5 100644
--- a/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.zh.md
+++ b/.agents/notes/implemented/process/2026-07-26-ci-failover-runbook.zh.md
@@ -6,16 +6,20 @@ Status: implemented
 
 ## 问题
 
-[CI](../../../../.github/workflows/ci.yml) 中三个必需的 Linux 工作作业（`node 24 / static`、`node 24 / coverage`、`node 24 / snapshots and artifacts`）运行在托管的企业级 32 核池上；聚合它们的必需判定作业（`all checks passed`）运行在标准 `ubuntu-latest` 上。当企业池发生故障——作业无限排队或企业标签消失——所有开启的拉取请求都无法合并，而"合并一个修复"这一常规恢复手段本身正被那些无法运行的必需检查死锁。**适用范围：本切换恢复的是企业级 Linux 池故障。**判定作业的其余必需依赖（`node-compat`、`python-sdk`、`windows`）按设计留在标准托管运行器上（可移植边界）；若更大范围的 GitHub 托管容量故障连标准池一并击倒，这些依赖仍会阻塞 `all checks passed`，且只有 Windows 这条腿完全没有自有替代——2026-07-27 的故障中标准池率先恢复，本设计押注的正是这一顺序。因此故障需要一个任何具备仓库写权限的响应者都能在不合并任何代码的情况下触发的开关。
+[CI](../../../../.github/workflows/ci.yml) 中三个必需的 Linux 工作作业（`node 24 / static`、`node 24 / coverage`、`node 24 / snapshots and artifacts`）运行在托管的企业级 32 核池上；聚合它们的必需判定作业（`all checks passed`）运行在标准 `ubuntu-latest` 上。必需的 Windows 作业（`windows node 24 / native complete`）运行在托管的企业级大型运行器上。当企业池发生故障——作业无限排队或企业标签消失——所有开启的拉取请求都无法合并，而"合并一个修复"这一常规恢复手段本身正被那些无法运行的必需检查死锁。**适用范围：本切换恢复的是企业级 Linux 池故障和托管 Windows 池故障。**判定作业的其余必需依赖（`node-compat`、`python-sdk`）按设计留在标准托管运行器上（可移植边界）；若更大范围的 GitHub 托管容量故障连标准池一并击倒，这些依赖仍会阻塞 `all checks passed`。因此故障需要一个任何具备仓库写权限的响应者都能在不合并任何代码的情况下触发的开关。
 
 ## 决策
 
-三个必需的 Linux 工作作业——以及 `all checks passed` 判定作业（若不随切换，即使全部工作作业通过，它仍会滞留在故障池的队列中）——各自通过仓库变量 `DSH_CI_FAILOVER` 解析运行器池。变量不存在（正常）时它们运行在托管企业池上；由任何具备写权限的协作者设为 `selfhosted` 时，四者全部切换到公司自有的自托管 `vm-backup` 池，覆盖率与快照的并发降到共享虚拟机上限，并跳过托管路径的 pnpm 缓存恢复。这个开关是写者可管理的仓库状态而非一次合并，因此在所有检查都是红色时仍然有效。自有池的就绪状态由 `serial / linux (self-hosted standby)` 通道持续验证——每次 master 推送都在其上运行完整的未分片聚合流程。
+三个必需的 Linux 工作作业、必需的 Windows 作业，以及 `all checks passed` 判定作业（若不随切换，即使全部工作作业通过，它仍会滞留在故障池的队列中）——各自通过仓库变量 `DSH_CI_FAILOVER` 解析运行器池。变量不存在（正常）时它们运行在托管企业池上；由任何具备写权限的协作者设为 `selfhosted` 时，五者全部切换到公司自有的自托管池：Linux 作业与判定位移入 `vm-backup` 池（`[self-hosted, linux, x64, vm-backup]`），覆盖率与快照的并发降到共享虚拟机上限，并跳过托管路径的 pnpm 缓存恢复；Windows 作业移入 `dsh-win-ci` 池（`[self-hosted, dsh-win-ci, windows]`）。这个开关是写者可管理的仓库状态而非一次合并，因此在所有检查都是红色时仍然有效。自有池的就绪状态由 `serial / linux (self-hosted standby)` 和 `serial / windows (self-hosted standby)` 通道持续验证——每次 master 推送都在其上运行完整的未分片聚合流程。
 
 ### 自有池是什么
 
 `vm-backup`：一台 64 核虚拟机，6 个常驻 systemd 管理的运行器实例。其镜像必须预装 Playwright Chromium 的 Linux 系统软件包；CI 会下载锁文件选定的浏览器，但绝不在这台持久化共享主机上运行 `apt`。切换前先看 `serial / linux (self-hosted standby)` 最近一次运行：其聚合流程包含浏览器回放，因此绿色热备同时验证常规容量和这项浏览器先决条件。
 
+#### Windows 池
+
+`dsh-win-ci`：公司自有 Windows CI 服务器上的 16 个常驻运行器实例。标签：`[self-hosted, dsh-win-ci, windows]`。镜像必须预装 Node 24、pnpm、Git，并启用开发人员模式以支持符号链接；Playwright Chromium 由 `serial / windows (self-hosted standby)` 通道在每次运行时安装。切换前先看 `serial / windows (self-hosted standby)` 最近一次运行：绿色热备验证该池可完整执行 `check:ci:windows-complete`。
+
 ### 切换步骤（任何具备写权限的协作者，约 1 分钟，无需合并）
 
 1. 仓库 **Settings → Secrets and variables → Actions → Variables → New repository variable**：名称 `DSH_CI_FAILOVER`，值 `selfhosted`。
diff --git a/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.i18n.yaml b/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.i18n.yaml
index c0a7dcfc02..b619bdc1d8 100644
--- a/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.i18n.yaml
+++ b/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write .agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md
-2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md: 499141ca6a3703a12d10c195b732441d49419599
-2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.zh.md: c57b7d32cd7204a845ab1884e9ae0b9045749a1c
+2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md: 612c42d19e3891f3fa036fea74b36e36da2fb5cd
+2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.zh.md: 7c82c787f11e93cbb62209d25df741d0abeba3dd
diff --git a/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md b/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md
index 499141ca6a..612c42d19e 100644
--- a/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md
+++ b/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.md
@@ -13,8 +13,8 @@ Outside `landlock-run.yml`, each workflow that installed pnpm hand-provisioned i
 `pnpm/action-setup@v4` is the only pnpm provisioning mechanism in CI: no workflow runs `corepack enable`. The root dev dependency on `@yarnpkg/cli-dist` separately supplies the modern Yarn CLI exercised by the generated-project e2e; package-manager coverage therefore does not inherit the runner image's Yarn Classic. Caching remains per-job policy on top of pnpm provisioning, in three deliberate shapes:
 
 - **Symmetric cache** (restore and save): `actions/setup-node` with `cache: pnpm` — `e2e.yml`, `docs-pages.yml`, `pi-ai-provider-e2e.yml`, `build-exe-for-python-sdk.yml`, and the node-compat and two benchmark jobs of `ci.yml`. The larger-runner benchmark keeps its store cache Linux-only through a conditional `cache:` input; the consolidated benchmark caches on both platforms.
-- **Restore-only / producer pairing** (hand-rolled `actions/cache` steps): the three enterprise-runner PR jobs and the Wine-based required Windows job restore without saving, keeping cache compression/upload off their latency-sensitive paths — an asymmetry `setup-node`'s cache cannot express. Each configures a store outside the action's replaceable install directory and resolves that path, matching the master-push serial-linux producer's path and exact key; the enterprise jobs skip restore during self-hosted failover because that VM's persistent store is already warm.
-- **Cache-less or persistent** (no store-cache action): the independent native Windows job, native serial-windows and serial-macos, plus `sandbox.yml` install from a cold or runner-local store. Extracting the many-file pnpm store costs more than a clean Windows install; the self-hosted standby and failover jobs instead reuse their VM's persistent pnpm store without transferring a hosted cache archive.
+- **Restore-only / producer pairing** (hand-rolled `actions/cache` steps): the three enterprise-runner PR jobs restore without saving, keeping cache compression/upload off their latency-sensitive paths — an asymmetry `setup-node`'s cache cannot express. Each configures a store outside the action's replaceable install directory and resolves that path, matching the master-push serial-linux producer's path and exact key; the enterprise jobs skip restore during self-hosted failover because that VM's persistent store is already warm.
+- **Cache-less or persistent** (no store-cache action): the required native Windows job, serial-windows standby, serial-macos, plus `sandbox.yml` install from a cold or runner-local store. Extracting the many-file pnpm store costs more than a clean Windows install; the self-hosted standby and failover jobs instead reuse their VM's persistent pnpm store without transferring a hosted cache archive.
 
 ## Alternatives considered
 
diff --git a/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.zh.md b/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.zh.md
index c57b7d32cd..7c82c787f1 100644
--- a/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.zh.md
+++ b/.agents/notes/implemented/process/2026-07-26-pnpm-action-setup-for-symmetric-ci-caching.zh.md
@@ -13,8 +13,8 @@ Status: implemented
 `pnpm/action-setup@v4` 是 CI 中提供 pnpm 的唯一机制：没有任何工作流运行 `corepack enable`。根目录的 `@yarnpkg/cli-dist` 开发依赖另行提供 generated-project e2e 所运行的现代 Yarn CLI（命令行界面）；因此，用于包管理器覆盖率的 Yarn 不会沿用 runner 镜像里的 Yarn Classic。缓存仍是叠加在 pnpm 提供机制上的按作业政策，保持三种刻意的形态：
 
 - **对称缓存**（既恢复也保存）：带 `cache: pnpm` 的 `actions/setup-node`——`e2e.yml`、`docs-pages.yml`、`pi-ai-provider-e2e.yml`、`build-exe-for-python-sdk.yml`，以及 `ci.yml` 的 node-compat 与两个 benchmark 作业。larger-runner benchmark 通过条件化的 `cache:` 输入让 store 缓存仅限 Linux；consolidated benchmark 在两个平台上都启用缓存。
-- **只恢复不上传／生产者配对**（手写的 `actions/cache` 步骤）：企业 runner 上的三个 PR（Pull Request）作业和基于 Wine 的必需 Windows 作业只恢复不保存，把缓存压缩／上传挡在它们的延迟敏感路径之外——这种不对称是 `setup-node` 的缓存无法表达的。每个作业都在 action 可替换的安装目录之外配置 store，并解析该路径，从而与 master 推送触发的 serial-linux 生产者所用的路径和精确键匹配；企业作业在自托管故障切换期间跳过恢复，因为该 VM 的持久 store 已经预热。
-- **无缓存或持久化**（不使用 store 缓存 action）：独立的原生 Windows 作业、原生 serial-windows 和 serial-macos，以及 `sandbox.yml` 均从冷 store 或 runner 本地 store 安装。解压含有大量文件的 pnpm store，成本高于在 Windows 上进行一次全新安装；自托管热备与故障切换作业则复用其 VM 的持久 pnpm store，不传输托管缓存归档。
+- **只恢复不上传／生产者配对**（手写的 `actions/cache` 步骤）：企业 runner 上的三个 PR（Pull Request）作业只恢复不保存，把缓存压缩／上传挡在它们的延迟敏感路径之外——这种不对称是 `setup-node` 的缓存无法表达的。每个作业都在 action 可替换的安装目录之外配置 store，并解析该路径，从而与 master 推送触发的 serial-linux 生产者所用的路径和精确键匹配；企业作业在自托管故障切换期间跳过恢复，因为该 VM 的持久 store 已经预热。
+- **无缓存或持久化**（不使用 store 缓存 action）：必需的 Windows 作业、serial-windows 热备、serial-macos，以及 `sandbox.yml` 均从冷 store 或 runner 本地 store 安装。解压含有大量文件的 pnpm store，成本高于在 Windows 上进行一次全新安装；自托管热备与故障切换作业则复用其 VM 的持久 pnpm store，不传输托管缓存归档。
 
 ## 曾考虑的替代方案
 
diff --git a/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml b/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml
deleted file mode 100644
index 07eb13b5cd..0000000000
--- a/.agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.i18n.yaml
+++ /dev/null
@@ -1,6 +0,0 @@
-# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
-# side as of the last confirmed-consistent state. Both languages carry equal authority;
-# after editing either side, bring the other along and re-record with:
-#   pnpm run verify-translation-pairing --write .agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.md
-2026-08-08-native-windows-pull-request-ci.md: 6a62fddb79670c3ab4cc0446796dbffd7130aed9
-2026-08-08-native-windows-pull-request-ci.zh.md: 990b8ed1434934337b8ff20c5f3be2c03cd9c61b
diff --git a/.github/AGENTS.md b/.github/AGENTS.md
index b2721fb84e..ad2aeac354 100644
--- a/.github/AGENTS.md
+++ b/.github/AGENTS.md
@@ -1,3 +1,3 @@
 # AGENTS.md — GitHub Actions
 
-Run jobs on Windows runners (`windows-*` labels) under native `pwsh`. The required pull-request `windows` job and the master `serial-windows` standby both run on real Windows; the self-hosted pool uses labels `[self-hosted, dsh-win-ci, windows]`. Under failover (`DSH_CI_FAILOVER=selfhosted`) the `windows` job retargets onto the self-hosted pool.
+Run Windows CI jobs under native `pwsh`. The required pull-request `windows` job and the master `serial-windows` standby both run on real Windows: normally on the hosted enterprise runner, failing over to the self-hosted `[self-hosted, dsh-win-ci, windows]` pool under `DSH_CI_FAILOVER=selfhosted`.
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 8a0482e103..f67a86c5fc 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -561,13 +561,6 @@ jobs:
         shell: pwsh
         run: pnpm install --frozen-lockfile
 
-      # The persistent VM image owns Playwright's Windows system dependencies;
-      # this step also proves that browser provisioning remains usable for
-      # failover.
-      - name: Install Playwright Chromium
-        shell: pwsh
-        run: pnpm --filter @deepseek-ai/dsh-frontend exec playwright install chromium
-
       - name: Run complete unsharded Windows gate inventory serially
         shell: pwsh
         env:
diff --git a/AGENTS.md b/AGENTS.md
index a8a202147c..7b326fbdb5 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -69,7 +69,6 @@ pnpm run typecheck
 pnpm run lint
 pnpm run duplication    # cross-file TypeScript clone detection
 pnpm run build          # tsc emits lib/types, tsdown bundles runtime
-pnpm run check:windows-wine  # ONLY when diagnosing a known Windows failure (needs wine); CI owns this signal
 pnpm run hygiene        # knip + publint + workspace constraints + NodeNext consumer check
 pnpm run doc-sync       # all documentation gates; leaf list in scripts/run-gates.ts
 pnpm run website:build  # VitePress build (doubles as dead-link check)
diff --git a/scripts/ci-workflow.spec.ts b/scripts/ci-workflow.spec.ts
index 2269c7cbee..b6d7d858ff 100644
--- a/scripts/ci-workflow.spec.ts
+++ b/scripts/ci-workflow.spec.ts
@@ -27,46 +27,60 @@ describe('CI workflow', () => {
     }
   })
 
-  it('keeps Wine blocking while native Windows reports independently', () => {
+  it('keeps a single required native Windows job with failover and a master-only standby', () => {
     const workflow = loadWorkflow('.github/workflows/ci.yml')
     if (!isRecord(workflow.jobs)
       || !isRecord(workflow.jobs.windows)
-      || !isRecord(workflow.jobs['windows-native'])
+      || !isRecord(workflow.jobs['serial-windows'])
       || !isRecord(workflow.jobs['all-checks-passed'])) {
-      throw new TypeError('CI workflow must define Wine, native Windows, and aggregate jobs')
+      throw new TypeError('CI workflow must define windows, serial-windows, and all-checks-passed jobs')
     }
 
     const windows = workflow.jobs.windows
-    const windowsNative = workflow.jobs['windows-native']
+    const serialWindows = workflow.jobs['serial-windows']
     const aggregate = workflow.jobs['all-checks-passed']
-    if (!Array.isArray(windows.steps) || !Array.isArray(windowsNative.steps) || !Array.isArray(aggregate.needs)) {
-      throw new TypeError('Windows jobs must define steps and the aggregate must define needs')
+    if (!Array.isArray(windows.steps) || !Array.isArray(aggregate.needs)) {
+      throw new TypeError('Windows job must define steps and the aggregate must define needs')
     }
-    const nativeCommandSteps = windowsNative.steps.filter((step): step is Record<string, unknown> & { run: string } => (
+    const commandSteps = windows.steps.filter((step): step is Record<string, unknown> & { run: string } => (
       isRecord(step) && typeof step.run === 'string'
     ))
 
-    expect(windows['runs-on']).toBe('ubuntu-latest')
-    expect(windows.name).toBe('windows node 24 / wine blocking')
+    // Required PR job: native Windows, failover-able, runs windows-complete.
+    expect(typeof windows['runs-on']).toBe('string')
+    expect(windows['runs-on']).toContain('DSH_CI_FAILOVER')
+    expect(windows['runs-on']).toContain('self-hosted')
+    expect(windows['runs-on']).toContain('dsh-win-ci')
+    expect(windows['runs-on']).toContain('dsh-windows-2025-16core')
+    expect(windows.name).toBe('windows node 24 / native complete')
     expect(windows.if).toBe("github.event_name == 'pull_request'")
-    expect(JSON.stringify(windows)).toContain('bash scripts/wine-windows-gates.sh')
-    expect(workflow.jobs).toHaveProperty('wine-apt-cache')
-    expect(windowsNative['runs-on']).toBe('dsh-windows-2025-16core')
-    expect(windowsNative.name).toBe('windows node 24 / native complete')
-    expect(windowsNative['timeout-minutes']).toBe(60)
-    expect(windowsNative.if).toBe("github.event_name == 'pull_request'")
-    expect(windowsNative.env).toMatchObject({
+    expect(windows.env).toMatchObject({
       DSH_COVERAGE_MAX_WORKERS: '2',
       DSH_GATE_CONCURRENCY: '2',
       DSH_PUBLINT_CONCURRENCY: '8',
     })
-    expect(windowsNative).not.toHaveProperty('continue-on-error')
-    expect(nativeCommandSteps).toHaveLength(3)
-    expect(nativeCommandSteps.every(step => step.shell === 'pwsh')).toBe(true)
-    expect(nativeCommandSteps.map(step => step.run)).toContain('pnpm run check:ci:windows-complete')
-    expect(JSON.stringify(windowsNative)).not.toMatch(/wine/i)
+    expect(commandSteps.every(step => step.shell === 'pwsh')).toBe(true)
+    expect(commandSteps.map(step => step.run)).toContain('pnpm run check:ci:windows-complete')
+    const developerMode = windows.steps.find((step): step is Record<string, unknown> & { run: string } => (
+      isRecord(step) && typeof step.run === 'string' && step.run.includes('AllowDevelopmentWithoutDevLicense')
+    ))
+    expect(developerMode).toBeDefined()
+
+    // No Wine or dual-lane remnants.
+    const workflowJson = JSON.stringify(workflow)
+    expect(workflowJson).not.toMatch(/wine/i)
+    expect(workflow.jobs).not.toHaveProperty('windows-native')
+    expect(workflow.jobs).not.toHaveProperty('wine-apt-cache')
+
+    // serial-windows: master-only standby, self-hosted, non-blocking.
+    expect(serialWindows.if).toBe("github.event_name == 'push' && github.ref == 'refs/heads/master'")
+    expect(serialWindows['runs-on']).toEqual(['self-hosted', 'dsh-win-ci', 'windows'])
+    expect(serialWindows.name).toBe('serial / windows (self-hosted standby)')
+
+    // Aggregate: windows required, serial-windows excluded.
     expect(aggregate.needs).toContain('windows')
     expect(aggregate.needs).not.toContain('windows-native')
+    expect(aggregate.needs).not.toContain('serial-windows')
   })
 
   it('keeps supported LSP source under native Windows coverage', () => {
diff --git a/scripts/wine-windows-gates.sh b/scripts/wine-windows-gates.sh
deleted file mode 100755
index 5e46b75f91..0000000000
--- a/scripts/wine-windows-gates.sh
+++ /dev/null
@@ -1,282 +0,0 @@
-#!/usr/bin/env bash
-# Run the blocking Windows gates (workspace build, production site) with real
-# win-x64 Node.js under Wine — the same script the pull-request `windows` job
-# in ci.yml executes and the optional local gate `pnpm run check:windows-wine`
-# wraps. Owning rationale and fidelity limits:
-# .agents/notes/implemented/process/2026-08-08-native-windows-pull-request-ci.md
-#
-# The working tree is never mutated: tracked plus untracked-unignored files
-# are snapshotted into a scratch directory, the Wine-specific pnpm overrides
-# (hoisted layout, win32-x64 platform packages) are appended to the SNAPSHOT's
-# pnpm-workspace.yaml, and the install and gates run there against the shared
-# pnpm store. The Wine prefix and the checksum-verified Windows Node zip
-# persist in .cache/wine-windows/ so reruns skip provisioning.
-#
-# Environment: DSH_WINE_NODE_MAJOR (default $PRIMARY_NODE_VERSION, then 24)
-# picks the Windows Node line; DSH_WINE_GATE_CACHE_DIR relocates the cache;
-# DSH_WINE_GATE_KEEP=1 preserves the scratch tree for inspection.
-
-set -euo pipefail
-
-repo_root="$(git rev-parse --show-toplevel)"
-node_major="${DSH_WINE_NODE_MAJOR:-${PRIMARY_NODE_VERSION:-24}}"
-cache_dir="${DSH_WINE_GATE_CACHE_DIR:-$repo_root/.cache/wine-windows}"
-
-export WINEDEBUG='-all'
-export WINEARCH=win64
-# Skip Wine Mono / Gecko installers: Node needs neither.
-export WINEDLLOVERRIDES='mscoree,mshtml='
-export WINEPREFIX="$cache_dir/prefix"
-
-# ---- preflight: fail loud before any expensive work --------------------
-wine_bin=''
-for candidate in "$(command -v wine || true)" "$(command -v wine64 || true)" /usr/lib/wine/wine64; do
-  if [ -n "$candidate" ] && [ -x "$candidate" ]; then wine_bin="$candidate"; break; fi
-done
-# GNU coreutils sha256sum on Linux; perl shasum ships with macOS. Both
-# accept the same "<hash>  <file>" --check input.
-checksum_tool=''
-if command -v sha256sum > /dev/null; then
-  checksum_tool='sha256sum'
-elif command -v shasum > /dev/null; then
-  checksum_tool='shasum'
-fi
-missing=()
-[ -n "$wine_bin" ] || missing+=('wine (apt: wine | brew: wine-stable)')
-command -v curl > /dev/null || missing+=('curl')
-command -v unzip > /dev/null || missing+=('unzip')
-[ -n "$checksum_tool" ] || missing+=('sha256sum or shasum (apt: coreutils | macOS ships shasum)')
-if ! command -v pnpm > /dev/null; then corepack enable > /dev/null 2>&1 || true; fi
-command -v pnpm > /dev/null || missing+=('pnpm (corepack enable)')
-if (( ${#missing[@]} > 0 )); then
-  printf 'wine-windows-gates: missing required tool: %s\n' "${missing[@]}" >&2
-  exit 1
-fi
-
-# Verify file $2 against SHA-256 hex $1 with whichever tool preflight found.
-verify_sha256() {
-  case "$checksum_tool" in
-    sha256sum) printf '%s  %s\n' "$1" "$2" | sha256sum --check - > /dev/null ;;
-    shasum) printf '%s  %s\n' "$1" "$2" | shasum -a 256 --check - > /dev/null ;;
-  esac
-}
-
-scratch="$(mktemp -d "${TMPDIR:-/tmp}/dsh-wine-gates.XXXXXX")"
-cleanup() {
-  wineserver -k > /dev/null 2>&1 || true
-  if [ "${DSH_WINE_GATE_KEEP:-0}" = '1' ]; then
-    echo "wine-windows-gates: scratch tree kept at $scratch"
-  else
-    rm -rf "$scratch"
-  fi
-}
-trap cleanup EXIT
-mkdir -p "$cache_dir" "$scratch/logs"
-
-# ---- provision Windows Node, boot Wine, snapshot + install concurrently ----
-curl_metadata_args=(
-  --fail --silent --show-error --location
-  --retry 3 --retry-all-errors --retry-delay 2
-  --http1.1 --connect-timeout 10 --max-time 30 --retry-max-time 120
-)
-
-download_node_archive() {
-  local version="$1" output="$2" attempt status=0
-  local archive="node-$version-win-x64.zip"
-  local primary_url="https://nodejs.org/dist/$version/$archive"
-  local mirror_url="https://npmmirror.com/mirrors/node/$version/$archive"
-
-  if curl --fail --silent --show-error --location --http1.1 \
-    --connect-timeout 10 --max-time 300 --speed-limit 1024 --speed-time 30 \
-    -o "$output" "$primary_url"; then
-    return 0
-  fi
-  echo 'wine-windows-gates: nodejs.org archive transfer stalled; resuming from the checksum-untrusted transport mirror' >&2
-  for attempt in 1 2 3; do
-    if curl --fail --silent --show-error --location --http1.1 \
-      --continue-at - --connect-timeout 10 --max-time 300 \
-      --speed-limit 1024 --speed-time 30 \
-      -o "$output" "$mirror_url"; then
-      return 0
-    else
-      status=$?
-    fi
-    (( attempt < 3 )) || break
-    echo "wine-windows-gates: mirror transfer failed (exit $status) on attempt $attempt; resuming partial download" >&2
-  done
-  return "$status"
-}
-
-provision_node() {
-  # Latest release of the primary line, checksum-verified against the same
-  # dist directory. Bound and retry every transfer so a stalled nodejs.org
-  # response cannot consume the entire CI job. Offline runs fall back to the
-  # newest cached zip, loudly.
-  local version zip
-  version="$(curl "${curl_metadata_args[@]}" https://nodejs.org/dist/index.json 2> /dev/null \
-    | node -e "let d='';process.stdin.on('data',c=>d+=c).on('end',()=>{const v=JSON.parse(d).find(r=>r.version.startsWith('v$node_major.'));if(v)console.log(v.version)})" \
-    || true)"
-  if [ -n "$version" ]; then
-    zip="$cache_dir/node-$version-win-x64.zip"
-    if [ ! -f "$zip" ]; then
-      download_node_archive "$version" "$zip.tmp"
-      local expected
-      expected="$(curl "${curl_metadata_args[@]}" "https://nodejs.org/dist/$version/SHASUMS256.txt" \
-        | awk -v a="node-$version-win-x64.zip" '$2 == a { print $1; exit }')"
-      [ -n "$expected" ] || { echo "wine-windows-gates: no SHASUMS256 entry for node-$version-win-x64.zip" >&2; exit 1; }
-      verify_sha256 "$expected" "$zip.tmp"
-      mv "$zip.tmp" "$zip"
-    fi
-  else
-    zip="$(ls -t "$cache_dir"/node-v"$node_major".*-win-x64.zip 2> /dev/null | head -1 || true)"
-    [ -n "$zip" ] || { echo "wine-windows-gates: nodejs.org unreachable and no cached Windows Node v$node_major zip in $cache_dir" >&2; exit 1; }
-    echo "wine-windows-gates: nodejs.org unreachable; using cached $(basename "$zip")" >&2
-  fi
-  unzip -q -o "$zip" -d "$scratch/node-win"
-  echo "$scratch/node-win/$(basename "$zip" .zip)/node.exe" > "$scratch/node-win-path"
-}
-
-boot_wine() {
-  "$wine_bin" wineboot --init > /dev/null 2>&1 || true
-  wineserver -w || true
-}
-
-snapshot_and_install() {
-  # Tracked + untracked-unignored files, minus agent-session litter; the
-  # existence filter drops paths staged as deleted. Then the Wine-specific
-  # install-time overrides go on the SNAPSHOT only: hoisted because Windows
-  # Node under Wine does not realpath pnpm's isolated-layout symlinks, and
-  # win32-x64 so the Windows esbuild/rolldown/rollup binaries materialize.
-  # Neither is recorded in the lockfile, so --frozen-lockfile stays valid;
-  # --ignore-scripts skips host lifecycle scripts no gate loads.
-  git -C "$repo_root" ls-files -z --cached --others --exclude-standard -- . ':!:.claude' ':!:.codex' \
-    | while IFS= read -r -d '' file; do [ -e "$repo_root/$file" ] && printf '%s\0' "$file"; done \
-    | tar -C "$repo_root" --null --files-from=- -cf - \
-    | tar -C "$scratch/tree" -xf -
-  cat >> "$scratch/tree/pnpm-workspace.yaml" << 'EOF'
-
-nodeLinker: hoisted
-supportedArchitectures:
-  os: [current, win32]
-  cpu: [current, x64]
-EOF
-  # The hoisted linker — used only by this lane — has an upstream rename
-  # race (pnpm/pnpm#12880): parallel linkers staging a nested package copy
-  # (observed on the tree's nested esbuild versions) rename their _tmp_*
-  # directory onto a path another racer already claimed, and the loser
-  # exits ERR_PNPM_ENOENT although an identical re-install succeeds.
-  # Exactly that signature earns up to two retries on a clean tree — the
-  # snapshot contains no node_modules, so wiping them restores the
-  # pre-install state; any other failure, or the race still standing after
-  # the final attempt, fails loud with the log tail.
-  local attempt
-  for attempt in 1 2 3; do
-    (cd "$scratch/tree" && pnpm install --frozen-lockfile --ignore-scripts > "$scratch/logs/install.log" 2>&1) \
-      && return 0
-    grep -q 'ERR_PNPM_ENOENT.*rename.*_tmp_' "$scratch/logs/install.log" || break
-    (( attempt < 3 )) || break
-    echo "wine-windows-gates: pnpm hoisted-linker rename race (pnpm/pnpm#12880) on install attempt $attempt; retrying on a clean tree" >&2
-    find "$scratch/tree" -name node_modules -type d -prune -exec rm -rf {} +
-  done
-  tail -40 "$scratch/logs/install.log" >&2
-  return 1
-}
-
-mkdir "$scratch/tree"
-start=$SECONDS
-provision_node & node_pid=$!
-boot_wine & wine_pid=$!
-snapshot_and_install & install_pid=$!
-# Wait for EVERY child before judging any: a bare `wait` under set -e would
-# exit on the first failure and let the EXIT trap delete $scratch while the
-# other children still run inside it. Named statuses also make the report
-# point at the root cause instead of a downstream symptom.
-node_status=0; wait "$node_pid" || node_status=$?
-wine_status=0; wait "$wine_pid" || wine_status=$?
-install_status=0; wait "$install_pid" || install_status=$?
-provision_failed=0
-report_provision() {
-  if (( $2 != 0 )); then
-    echo "wine-windows-gates: FAILED $1 (exit $2)" >&2
-    provision_failed=$2
-  fi
-}
-report_provision 'Windows Node provisioning' "$node_status"
-report_provision 'wineboot' "$wine_status"
-report_provision 'workspace snapshot + pnpm install' "$install_status"
-if (( provision_failed != 0 )); then exit "$provision_failed"; fi
-node_win="$(cat "$scratch/node-win-path")"
-echo "wine-windows-gates: provisioned in $((SECONDS - start))s (wine $("$wine_bin" --version 2> /dev/null), node $(basename "$(dirname "$node_win")"))"
-
-# ---- resolve entrypoints, lay the vue link, smoke ------------------------
-# Node under Wine cannot attach stdio to pipes the caller owns (Socket open
-# EBADF at bootstrap), so every invocation routes stdio through a file.
-wine_node() {
-  local log="$1"
-  shift
-  local status=0
-  "$wine_bin" "$node_win" "$@" < /dev/null > "$log" 2>&1 || status=$?
-  return "$status"
-}
-
-cd "$scratch/tree"
-tsc_js='node_modules/typescript/bin/tsc'
-tsdown_js='node_modules/tsdown/dist/run.mjs'
-vitepress_js='node_modules/vitepress/bin/vitepress.js'
-[ -f "$vitepress_js" ] || vitepress_js='website/node_modules/vitepress/bin/vitepress.js'
-for entry in "$tsc_js" "$tsdown_js" "$vitepress_js"; do
-  [ -f "$entry" ] || { echo "wine-windows-gates: expected entrypoint missing after hoisted install: $entry" >&2; exit 1; }
-done
-# VitePress links vue into the site's node_modules at build time; Wine cannot
-# CREATE Windows symlinks (ENOTSUP) but follows pre-existing Unix ones.
-if [ -d node_modules/vue ] && [ ! -e website/node_modules/vue ]; then
-  mkdir -p website/node_modules
-  ln -s ../../node_modules/vue website/node_modules/vue
-fi
-
-wine_node "$scratch/logs/smoke.log" -p "'smoke: ' + process.platform + ' ' + process.arch + ' ' + process.version"
-cat "$scratch/logs/smoke.log"
-grep -q '^smoke: win32 x64' "$scratch/logs/smoke.log" || { echo 'wine-windows-gates: Windows Node smoke did not report win32 x64' >&2; exit 1; }
-
-# ---- the two blocking surfaces, concurrently ------------------------------
-# The build preserves the face order from package.json: compile and bundle the
-# Host face before compiling and bundling the Client face.
-# Both statuses are captured so one failure cannot hide the other's result.
-build_gate() {
-  wine_node "$scratch/logs/host-tsc.log" "$tsc_js" -b tsconfig.host.json --pretty false || return $?
-  wine_node "$scratch/logs/host-tsdown.log" "$tsdown_js" --env.DSH_BUILD_FACE host || return $?
-  wine_node "$scratch/logs/client-tsc.log" "$tsc_js" -b tsconfig.client.json --pretty false || return $?
-  wine_node "$scratch/logs/client-tsdown.log" "$tsdown_js" --env.DSH_BUILD_FACE client
-}
-site_gate() {
-  cd website
-  wine_node "$scratch/logs/site.log" "../$vitepress_js" build .
-}
-
-start=$SECONDS
-build_gate & build_pid=$!
-site_gate & site_pid=$!
-build_status=0
-wait "$build_pid" || build_status=$?
-site_status=0
-wait "$site_pid" || site_status=$?
-elapsed=$((SECONDS - start))
-
-report() {
-  local label="$1" status="$2"
-  shift 2
-  if (( status == 0 )); then
-    echo "wine-windows-gates: PASS $label (${elapsed}s window)"
-  else
-    echo "== FAILED $label (exit $status) ==" >&2
-    for log in "$@"; do tail -n 200 "$log" >&2 || true; done
-  fi
-}
-report 'build (Host tsc/tsdown, Client tsc/tsdown)' "$build_status" \
-  "$scratch/logs/host-tsc.log" \
-  "$scratch/logs/host-tsdown.log" \
-  "$scratch/logs/client-tsc.log" \
-  "$scratch/logs/client-tsdown.log"
-report 'production site (vitepress build)' "$site_status" "$scratch/logs/site.log"
-if (( build_status != 0 )); then exit "$build_status"; fi
-exit "$site_status"