Merge latest master into PR 555

# Conflicts:
#	docs/architecture.i18n.yaml
#	docs/core-data-structures/core.i18n.yaml
#	examples/acp-agent/tests/snapshots/cordis-inspect-jsdoc/session.jsonl
#	packages/client/runtime/README.i18n.yaml
#	packages/client/ui-conversation/README.i18n.yaml
#	packages/client/ui-conversation/src/client/chat/MessageItem.tsx
#	packages/compact/compact-basic/README.i18n.yaml
#	packages/ui/tui/README.i18n.yaml

.agents/notes/implemented/architecture/2026-06-21-bounded-llm-request-recovery.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/architecture/2026-06-21-bounded-llm-request-recovery.md
-2026-06-21-bounded-llm-request-recovery.md: 5c76ed5d754ea40f41dff78cb56ee7fc139a32b1
-2026-06-21-bounded-llm-request-recovery.zh.md: 1fa56f3fe0405cab663c2843d423a78d910170dd
+2026-06-21-bounded-llm-request-recovery.md: 24725dcf300cf69e9cc72580d0c8afe937d4e2b9
+2026-06-21-bounded-llm-request-recovery.zh.md: 5f03a65b00be8d3349addce82e4f3faa2af1fe7e

.agents/notes/implemented/architecture/2026-06-21-bounded-llm-request-recovery.md

@@ -84,7 +84,7 @@ Boundary tests prove termination at both actual transports. The hand-written ada
 
 A failed attempt may leave `assistant/chunk` events in its closed step, but it never appends `assistant/message` and never dispatches a tool. A retry closes the failed turn, opens the next numbered turn, reconstructs the request from the durable surface, and produces its own chunks. UIs may render live chunks while a step is open, then mark or clear that transient view when `llm/retry` identifies the failed step or `turn/end` records failure. Web validates the complete retry payload contract, clears the failed partial at `llm/retry`, projects consecutive retry-turn events into one stable row updated to the latest attempt, and derives scheduled, started, or cancelled status from subsequent turn facts. Its countdown anchors the scheduled delay to browser receipt rather than the Host event clock, uses ceiling-rounded seconds with a one-second floor, animates only while unresolved, and keeps exact latest failure details collapsed behind the row. Retry nodes anchor their own trajectory turn even when the failed attempt has no assistant node. Message derivation continues to ignore the failed chunks, and Web applies the same projection during history rebuild so refreshing cannot resurrect discarded partials or duplicate retry rows.
 
-If recovery is exhausted, the final failure is stored once on `turn/end.reason` with the structured facts. If transient recovery continues, `llm/retry` is the durable home for that attempt's failure and delay. No standalone final-error event or response-id vocabulary is added.
+If recovery is exhausted, the final failure is stored once on `turn/end.reason` with the structured facts. Web derives one `turn-error` node at that sequence position and renders its display-safe message and optional code inline; AUTH projections replace provider copy that may echo credential fragments with `API key is invalid`, while the raw diagnostic remains in the session log. The same fold runs for live events and history replay. If transient recovery continues, `llm/retry` is the durable home for that attempt's failure and delay, so its failed turn does not also gain a terminal error row. No standalone final-error event or response-id vocabulary is added.
 
 ## Out of scope
 

.agents/notes/implemented/architecture/2026-06-21-bounded-llm-request-recovery.zh.md

@@ -84,7 +84,7 @@ agent-spine 演示组合包加载该插件，因此共享的 stdio/TUI、一次
 
 一次失败尝试可以在已关闭的步骤中留下 `assistant/chunk` 事件，但绝不会追加 `assistant/message`，也不会分发工具。重试会关闭失败轮次，开启下一个编号轮次，从持久表层重建请求，并生成自己的分片。步骤仍处于打开状态时，UI 可以渲染实时分片；当 `llm/retry` 标识失败步骤，或 `turn/end` 记录失败时，UI 再标记或清除这份暂时视图。Web 会验证完整的重试载荷契约，在 `llm/retry` 到达时清除失败的部分输出，将连续重试轮次的事件投影为稳定的一行，并用最新一次尝试更新该行，再从后续轮次事实派生 scheduled、started 或 cancelled 状态。倒计时以浏览器收到事件的时刻为计划延迟的起点，而不是使用 Host 事件时钟；它按向上取整且不低于 1 秒的秒数显示，仅在重试尚未结束时显示动画，并把最近一次失败的准确详情折叠在该行之后。即使失败尝试没有 assistant 节点，重试节点也会锚定自身的轨迹轮次。消息派生仍会忽略失败分片；Web 在重建历史时也会应用同一投影，因此刷新页面不会让已丢弃的部分输出重新出现，也不会生成重复的重试行。
 
-如果恢复预算耗尽，最终失败会连同结构化事实在 `turn/end.reason` 中存储一次。如果暂时性恢复继续，`llm/retry` 就是该次尝试的失败与延迟的持久归属位置。本决策不增加独立的最终错误事件或响应 id 词汇。
+如果恢复预算耗尽，最终失败会连同结构化事实在 `turn/end.reason` 中存储一次。Web 会在该序列位置派生一个 `turn-error` 节点，并内联渲染适合展示的消息与可选错误码；AUTH 投影会把可能回显凭据片段的提供方文案替换为 `API key is invalid`，原始诊断仍保留在会话日志中。实时事件和历史回放使用同一套折叠逻辑。如果暂时性恢复继续，`llm/retry` 就是该次尝试的失败与延迟的持久归属位置，因此该失败轮次不会再获得终态错误行。本决策不增加独立的最终错误事件或响应 id 词汇。
 
 ## 不在范围内
 

.agents/notes/implemented/architecture/2026-07-30-session-end-seed-log-boundary.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/architecture/2026-07-30-session-end-seed-log-boundary.md
-2026-07-30-session-end-seed-log-boundary.md: 268646e192d0b8e0a5dde03957a18ef155b7038e
-2026-07-30-session-end-seed-log-boundary.zh.md: dca87e16de5e567ff85d2b32b8243f76ebed1c4a
+2026-07-30-session-end-seed-log-boundary.md: ca6145f3ce404b88d2a144483c6f1145fe0a9a7e
+2026-07-30-session-end-seed-log-boundary.zh.md: d492c4ac5f39328e75d991b285004e2d19c56283

.agents/notes/implemented/architecture/2026-07-30-session-end-seed-log-boundary.md

@@ -52,4 +52,4 @@ Cost: a seeded session's log is one event longer, including an empty resumed log
 
 `session/end-seed` joins the on-disk vocabulary. Under the pre-release stance (`SESSION_FORMAT_VERSION` pinned at `0`, no compatibility promise) older logs simply lack it, and a log without a boundary correctly classifies nothing as constructor-seed history.
 
-Not built here: no plugin reads the boundary yet. Wiring the compaction seam's staleness check to it is the follow-up that motivated this boundary; the predicate helper belongs with that seam, where a real consumer decides its shape, rather than shipping into core untested against one.
+The [queued manual compaction decision](../feature/2026-07-30-queued-manual-compaction.md) now supplies the first consumer. Its tail scan independently finds the unmatched `compact/start` and newest end-seed, treats only a start after that boundary as live, and clears the invariant trace on the same replay transition. The predicate remains in the compaction package rather than becoming a generic core helper.

.agents/notes/implemented/architecture/2026-07-30-session-end-seed-log-boundary.zh.md

@@ -52,4 +52,4 @@ Status: implemented
 
 `session/end-seed` 加入了落盘词汇表。在预发布立场下（`SESSION_FORMAT_VERSION` 固定为 `0`，不作兼容承诺），更旧的日志只是没有它，而没有边界的日志会正确地判定没有任何内容属于构造种子历史。
 
-此处未做：还没有任何插件读取该边界。把压缩 seam 的陈旧性检查接到它上面，是催生这条边界的后续工作；谓词辅助函数应当归属那个 seam——在那里由真实消费方决定它的形状——而不是未经真实消费方检验就先落进核心。
+[排队手动压缩决策](../feature/2026-07-30-queued-manual-compaction.md)如今提供了第一个消费方。其尾部扫描会分别查找未匹配的 `compact/start` 与最新 end-seed，只把位于该边界之后的 start 视为活动锁，并在同一个回放转换上清除不变量追踪状态。该谓词仍位于压缩包中，不会成为通用核心辅助函数。

.agents/notes/implemented/bug-fix/2026-07-29-human-transcript-append-origin.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/bug-fix/2026-07-29-human-transcript-append-origin.md
-2026-07-29-human-transcript-append-origin.md: dcc4a786c6f1926f06dce03124ec1d8ca805d7ae
-2026-07-29-human-transcript-append-origin.zh.md: 0fefc52afa52e99cdec2bcea1a86b9c28711dd67
+2026-07-29-human-transcript-append-origin.md: a47dd49dd831cdd32d520137417bf47d2c056a09
+2026-07-29-human-transcript-append-origin.zh.md: 31639bd9aac5d6dace80392004f37c747bff2c36

.agents/notes/implemented/bug-fix/2026-07-29-human-transcript-append-origin.md

@@ -26,7 +26,7 @@ No persisted event, RPC envelope, compaction transaction, or model-visible surfa
 
 The browser client is fixed separately, in [the web transcript projection note](2026-07-30-web-transcript-log-ordered-projection.md): it projects the same append-origin transcript in log order and renders a marker component, and it closes the pagination hole this change opened — because `session.history` no longer spends quota on the checkpoint, it never cuts on the checkpoint's provenance group, so a page can carry a checkpoint citing a `surfaceOp.start` outside the window, which the browser's surface fold rejected. That hole predates this change (counting could already run past a checkpoint into the range it shadows), but the old rule accidentally covered the case where the checkpoint was the oldest counted message and pulled the whole shadowed range onto its page.
 
-Rendering compaction *progress* — a terminal indicator while a compaction runs — needs the bracket-first ordering that the queued manual `/compact` work introduces, and is out of scope here. The marker also carries no scale: the checkpoint's `sourceEventSeqs` already hold the shadowed count, so a count or range would tell a reader how much each row folded. That belongs with progress, where the reader meets the other half of the same information. Whoever takes it should fold the terminal's two replacement branches — replay and the live listener, textually identical and 600 lines apart — into one `renderReplacement(event)` first, so the marker's content has a single home.
+The terminal's [live compaction progress decision](../feature/2026-07-30-compaction-progress-visibility.md) uses standalone bracket events to drive the existing one-cell indicator. It does not change the completion marker owned here or add scale: the checkpoint's `sourceEventSeqs` remain available for a separately justified count or range. Progress therefore needs neither marker-content changes nor a prerequisite `renderReplacement(event)` extraction.
 
 ## Alternatives considered
 

.agents/notes/implemented/bug-fix/2026-07-29-human-transcript-append-origin.zh.md

@@ -26,7 +26,7 @@ Status: implemented
 
 浏览器客户端在[Web 记录投影笔记](2026-07-30-web-transcript-log-ordered-projection.md)中单独修复：它按日志顺序投影同一份 append 来源记录并渲染一个标记组件，同时闭合本次变更打开的分页缺口——因为 `session.history` 不再为检查点消耗额度，它永远不会按检查点的溯源分组切分，于是一页可以携带一个引用了窗口之外 `surfaceOp.start` 的检查点，而浏览器的 surface fold 会拒绝该范围。这个缺口早于本次变更（此前计数就可能越过检查点进入它所遮蔽的范围），但旧规则恰好覆盖了这样一种情形：检查点是最旧的被计数消息，其溯源分组把整段被遮蔽的范围一起拉到该页。
 
-渲染压缩*进度*——压缩运行期间的终端指示——需要排队式手动 `/compact` 工作引入的“先开括号”顺序，不在本次范围内。标记同样不携带规模信息：检查点的 `sourceEventSeqs` 已经包含被遮蔽的数量，因此一个计数或区间可以告诉读者每一行折叠了多少内容。这件事属于进度那一侧，读者正是在那里遇到同一份信息的另一半。接手者应当先把终端里两处替换分支——回放与实时监听器，文本完全相同却相隔 600 行——合并为一个 `renderReplacement(event)`，让标记的内容只有一个归处。
+终端的[实时压缩进度决策](../feature/2026-07-30-compaction-progress-visibility.md)使用独立标记对中的事件驱动现有的单格指示器。它既不改变本文所负责的完成标记，也不添加规模信息：检查点的 `sourceEventSeqs` 仍可供经另行论证的计数或区间使用。因此，进度显示既不需要修改标记内容，也不以提取 `renderReplacement(event)` 为前置条件。
 
 ## Alternatives considered
 

.agents/notes/implemented/bug-fix/2026-07-30-web-transcript-log-ordered-projection.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/bug-fix/2026-07-30-web-transcript-log-ordered-projection.md
-2026-07-30-web-transcript-log-ordered-projection.md: 9fb338643774efaeb9deab6f66920a9f4276ce67
-2026-07-30-web-transcript-log-ordered-projection.zh.md: 7962dd432b8bbf115acde9dd480eba9c91f35bfc
+2026-07-30-web-transcript-log-ordered-projection.md: 4bc629bfa619a2858ec0335a6136d3a5a24b025a
+2026-07-30-web-transcript-log-ordered-projection.zh.md: d011ce9453bc1dbb9fdb62372aa1126c95ffb143

.agents/notes/implemented/bug-fix/2026-07-30-web-transcript-log-ordered-projection.md

@@ -69,4 +69,4 @@ The web e2e scenario now seeds a real compaction transaction over its recorded t
 
 ## Deferred
 
-Compaction **progress** — an indicator while a compaction runs — needs the bracket-first ordering the queued manual-compaction work introduces, and stays out of scope here as it did in the terminal. The marker also carries no **scale**: the checkpoint's `sourceEventSeqs` already hold the shadowed count, so a count or range would tell a reader how much each row folded. Both belong together, where the reader meets the two halves of the same information.
+The terminal's [compaction progress decision](../feature/2026-07-30-compaction-progress-visibility.md) uses the live standalone bracket to drive a one-cell indicator and does not change this browser projection. The marker still carries no **scale**: the checkpoint's `sourceEventSeqs` hold the shadowed count, so a separately justified count or range can be added without coupling it to progress.

.agents/notes/implemented/bug-fix/2026-07-30-web-transcript-log-ordered-projection.zh.md

@@ -69,4 +69,4 @@ Web e2e 场景现在在它录制的那一轮之上播种一次真实的压缩事
 
 ## Deferred
 
-压缩**进度**——压缩运行期间的指示——需要排队式手动压缩工作引入的“先开括号”顺序，与终端一样不在本次范围内。标记同样不携带**规模**信息：检查点的 `sourceEventSeqs` 已经包含被遮蔽的数量，因此一个计数或区间可以告诉读者每一行折叠了多少内容。两者应当放在一起，读者正是在那里遇到同一份信息的两半。
+终端的[压缩进度决策](../feature/2026-07-30-compaction-progress-visibility.md)使用实时独立标记对驱动单格指示器，并不改变此浏览器投影。标记仍不携带**规模**信息：检查点的 `sourceEventSeqs` 保存被遮蔽的数量，因此可以另行论证后添加计数或区间，而无须将其与进度耦合。

.agents/notes/implemented/bug-fix/2026-07-31-hero-visible-while-blank-session-opens.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/implemented/bug-fix/2026-07-31-hero-visible-while-blank-session-opens.md
+2026-07-31-hero-visible-while-blank-session-opens.md: b39963beffa403ef6fa44735aa88395a99139751
+2026-07-31-hero-visible-while-blank-session-opens.zh.md: b451d7c00ec7eb8d736134e738d72e5e07fd1b04

.agents/notes/implemented/bug-fix/2026-07-31-hero-visible-while-blank-session-opens.md

@@ -0,0 +1,35 @@
+# Agent Note: Hero stays visible while a blank session opens
+
+Status: implemented
+
+English | [中文](2026-07-31-hero-visible-while-blank-session-opens.zh.md)
+
+## Problem
+
+The conversation root has a `settling` phase for a session that is still opening while its composer reads `blank`: the hero-versus-docked outcome is unknowable until history arrives, so the composer seat is hidden (`visibility:hidden`) rather than flashing the centered hero and snapping to the docked bar. Startup auto-selection turned that guard into the defect it was meant to prevent. From the no-workspace hero, `WorkspacesService.startInitialSelection` connects the most recent workspace and opens its blank session; `openState` flips to `loading` the moment `open()` lands, so the center column went blank for the whole history round-trip and then repainted, which reads as a full-page refresh on every launch.
+
+## Decision
+
+`ConversationRoot` reads the session list summary's `blank` flag alongside the conversation snapshot and exempts summary-proven blank sessions from settling: `settling` additionally requires `summaryBlank !== true`, and `hero` accepts a blank composer whenever the summary proves the session blank, in every open state rather than only `loading`. A session the list already reports as blank can only land on the hero, so hiding buys nothing and costs the visible flash; the same proof holds before the open starts (`cold`) and after one fails (`error`), where the previous conditions fell through to the active phase and rendered a docked bare composer under chrome `ConversationSession` hides for blank sessions. Whenever the summary does not prove the session blank — a row reporting `blank: false`, or no row at all because the list has not caught up — `summaryBlank` is not `true` and the conservative settling hide is unchanged.
+
+The summary flag and the snapshot's own `blank` are distinct sources: the snapshot describes the session being opened, the summary is the list row that already exists before the open resolves. Only the latter is available early enough to decide the phase.
+
+## Alternatives considered
+
+**Drop the settling phase entirely.** Rejected because it still earns its keep for a session with no summary row: without a prior claim about emptiness, hero-versus-docked is genuinely unknowable and the flash it prevents is the worse one.
+
+**Delay the `loading` flip until history returns.** Rejected because `openState` is authoritative about the open operation; deferring it to suppress a presentation artifact would misreport the data state to every other consumer.
+
+**Cross-fade or otherwise animate the settling hide.** Rejected because the column has nothing to show during the round-trip either way — the fix is to not hide content whose outcome is already known, not to decorate the hiding.
+
+## Deferred
+
+The no-session→session tree relocation in `ConversationRoot` (the hero/composer subtree moves into the `conversation.session` outlet) still rebuilds the composer DOM on the same transition; removing it means moving `conversation.session` to `session-maybe` scope, a slot-contract change that needs its own proposal.
+
+Object-layer reference churn found while diagnosing this — no-op projections minting fresh snapshots, the create path projecting twice, `select()` using `notifyNow` from async continuations — is real but independent of the visible flash.
+
+## Consequences
+
+Startup auto-selection renders the hero immediately and keeps the composer seat and header visible through the history round-trip, so launching into a recent workspace no longer looks like a page reload. Sessions whose summary does not prove them blank keep the previous settling behavior, so the guard still covers the case it was written for. Skeleton tests pin all three summary shapes: a row reporting `blank: false` settles, an absent row settles, and a summary-proven blank session opening under `loading` renders hero chrome with a live textarea.
+
+The assembled coverage is `apps/web/tests/startup-auto-selection.e2e.ts` (keyless web browser lane): it registers a workspace, holds the `session.history` response open at the browser's network boundary, and asserts the visible frame while the auto-selected open is in flight — hero phase, hero title, painted composer — plus a recorded phase timeline of exactly `['hero']` for the whole load. Holding the round-trip is what makes it a regression test rather than a race: against a loopback host the open settles too fast to sample, and with the exemption reverted the held window is precisely when the root reports `settling`.

.agents/notes/implemented/bug-fix/2026-07-31-hero-visible-while-blank-session-opens.zh.md

@@ -0,0 +1,35 @@
+# Agent Note: 空白会话打开期间保持 hero 可见
+
+Status: implemented
+
+[English](2026-07-31-hero-visible-while-blank-session-opens.md) | 中文
+
+## 问题
+
+会话根节点为"正在打开且 composer 处于 `blank`"的会话保留了一个 `settling` 阶段：在历史记录返回之前，hero 与 docked 的归属不可知，因此宁可隐藏 composer 座位（`visibility:hidden`），也不要先闪出居中的 hero 再跳到底部输入条。启动时的自动选择把这道防护变成了它本要防止的缺陷。从无工作区的 hero 进入时，`WorkspacesService.startInitialSelection` 会连接最近的工作区并打开其空白会话；`open()` 一落地 `openState` 立即翻为 `loading`，中间栏因此在整个历史往返期间保持空白，随后重绘一次——每次启动看起来都像整页刷新。
+
+## 决策
+
+`ConversationRoot` 在读取会话快照的同时读取会话列表摘要的 `blank` 标志，并让"摘要已证明为空白"的会话豁免 settling：`settling` 额外要求 `summaryBlank !== true`，而 `hero` 在摘要证明会话为空白时接受处于 blank 的 composer——覆盖全部 open state，而非仅 `loading`。列表已报告为空白的会话只可能落到 hero，因此隐藏毫无收益，只换来一次可见闪烁；同一份证明在打开开始之前（`cold`）与打开失败之后（`error`）同样成立，而此前的条件会在这两种状态下落到 active 阶段，在 `ConversationSession` 为空白会话隐藏的外壳之下渲染出一条停靠的裸 composer。只要摘要没有证明会话为空白——无论是报告 `blank: false` 的行，还是列表尚未跟上因而根本没有该行——`summaryBlank` 都不为 `true`，保守的 settling 隐藏行为保持不变。
+
+摘要标志与快照自身的 `blank` 是两个不同来源：快照描述正在打开的这个会话，摘要则是在打开操作完成之前就已存在的列表行。只有后者足够早，可用于决定阶段。
+
+## 备选方案
+
+**彻底移除 settling 阶段。** 否决，因为对没有摘要行的会话它仍有价值：在缺少任何关于"是否为空"的先验断言时，hero 与 docked 的归属确实不可知，而它所防止的那种闪烁更糟糕。
+
+**推迟 `loading` 的翻转，直到历史返回。** 否决，因为 `openState` 是打开操作的权威状态；为了压制一个呈现层瑕疵而推迟它，会向其他所有消费者误报数据状态。
+
+**为 settling 的隐藏加交叉淡入或其他动画。** 否决，因为无论如何该栏在往返期间都没有内容可展示——正确的修复是不隐藏结局已知的内容，而不是把隐藏装饰得好看些。
+
+## 推迟事项
+
+`ConversationRoot` 中"无会话→有会话"的树位置迁移（hero/composer 子树移入 `conversation.session` 出口）仍会在同一次转换中重建 composer 的 DOM；消除它意味着把 `conversation.session` 移到 `session-maybe` 作用域，这是一次插槽契约变更，需要单独立项。
+
+诊断期间发现的对象层引用抖动——空操作投影铸造出新的快照、创建路径重复投影一次、`select()` 在异步续体中使用 `notifyNow`——确实存在，但与这次可见闪烁相互独立。
+
+## 影响
+
+启动自动选择会立即渲染 hero，并在整个历史往返期间保持 composer 座位与 header 可见，因此启动进入最近工作区不再像页面重载。摘要未证明为空白的会话保持原有的 settling 行为，这道防护仍覆盖它当初针对的场景。骨架测试固定了摘要的三种形态：报告 `blank: false` 的行进入 settling；根本没有该行同样进入 settling；摘要已证明为空白的会话在 `loading` 期间渲染 hero 外壳与可用的文本框。
+
+组装级覆盖是 `apps/web/tests/startup-auto-selection.e2e.ts`（无密钥的 Web 浏览器泳道）：它注册一个工作区，在浏览器网络边界上扣住 `session.history` 的响应，并在自动选择的打开仍在飞行途中断言可见画面——hero 阶段、hero 标题、已绘制的 composer——外加整次加载记录到的阶段时间线恰好为 `['hero']`。扣住这次往返正是它成为回归测试而非竞态的原因：对着回环主机，打开会快到无从采样；而一旦回退这条豁免，被扣住的这段窗口恰恰就是根节点报告 `settling` 的时刻。

.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.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/feature/2026-06-18-compaction-capability-seam.md
-2026-06-18-compaction-capability-seam.md: 3c219b734e148b963fb5857de89c16f28c2bd402
-2026-06-18-compaction-capability-seam.zh.md: b2c7e9720b596705b60a284e6ccf1448a782b7fc
+2026-06-18-compaction-capability-seam.md: ef37313bc6fb984689793fa5a3e7ac4d9238ea88
+2026-06-18-compaction-capability-seam.zh.md: 9f123a8c40f303a2635af78cafd34de448e27e03

.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.md

@@ -18,10 +18,10 @@ Two forces shape the design. First, compaction policy and reusable token measure
 
 Per the [capability-seams Agent Note](../architecture/2026-06-13-capability-seams.md), compaction ships as separate packages so the contract, the algorithm, and (later) the consumer surface evolve independently:
 
-1. **Interface** — `@deepseek-ai/dsh-compact`: an abstract `CompactService` owning the `ctx.compact` key, the `CompactionResult` vocabulary, the `compact/*` session events, and the canonical checkpoint message source. It declares `compactIfNeeded()` and `compactRegion()` as **abstract** — the contract states *what* compaction does, not *how*.
+1. **Interface** — `@deepseek-ai/dsh-compact`: an abstract `CompactService` owning the `ctx.compact` key, the `CompactionResult` vocabulary, the `compact/*` session events, the manual failure taxonomy, and the canonical checkpoint message source. It declares `compactIfNeeded()`, `compactNow()`, and `compactRegion()` as **abstract** — the contract states *what* compaction does, not *how*.
 2. **Implementation** — `@deepseek-ai/dsh-compact-basic`: a concrete `BasicCompactService` that consumes `ctx.tokenMeter` and owns the tail→head retention walk, summarization via `ctx.llm.stream()`, the surface replacement, the lock, post-step pressure, and canonical context-overflow recovery. `summarize()` is its sole subclass hook; pricing and replay stay with the meter.
 3. **Model-free companion** — `@deepseek-ai/dsh-compact-tool-result-prune`: a concrete optional service that rewrites oversized current `tool/result` nodes before the backend selects a summary range. It is not a second compaction implementation and does not implement `CompactService`.
-4. **Consumer** — deferred. A `/compact` tool and slash command will `inject: ['compact']` and call the contract; they are intentionally out of scope here so the seam settles first.
+4. **Human consumer** — `@deepseek-ai/dsh-command-compact` registers argument-free `/compact` through `ctx.commands` and calls the backend-independent `compactNow()` operation. It is direct human control, not a model-facing tool.
 
 ### The contract depends on `dsh-session` and `dsh-llm` — a deliberate deviation
 
@@ -29,11 +29,11 @@ The capability-seams Agent Note states the interface package "depends only on co
 
 This is not a coupling smell — it is the contract's domain. The "only cordis" guidance was always shorthand for "the interface depends only on what the contract genuinely names, and never on an implementation." `dsh-session` and `dsh-llm` are themselves interface/vocabulary packages, not implementations; `dsh-compact` still imports no backend. The seam's real invariant — *consumers and implementations evolve independently behind an abstract service* — holds intact.
 
-### Abstract `compactIfNeeded` / `compactRegion`, algorithm in the backend
+### Three abstract operations, algorithm in the backend
 
-An earlier draft put the full algorithm (the retention walk, token-summing, text extraction) as concrete methods on the interface. That recouples the contract to one strategy: a backend that wants a different retention policy or event sequence would have to fight inherited concrete code. Making both core methods abstract puts every *how* decision in the backend and keeps the interface a statement of *what*. Token measurement is not a compaction hook at all; the singleton service lets multiple consumers share one per-session replay fold.
+An earlier draft put the full algorithm (the retention walk, token-summing, text extraction) as concrete methods on the interface. That recouples the contract to one strategy: a backend that wants a different retention policy or event sequence would have to fight inherited concrete code. Making all three operations abstract puts every *how* decision in the backend and keeps the interface a statement of *what*. Token measurement is not a compaction hook at all; the singleton service lets multiple consumers share one per-session replay fold.
 
-`compactIfNeeded(agent, trigger, signal)` takes an explicit `'pressure' | 'context-overflow'` trigger and cancellation. It reads only the latest durable routed request; no header means no work, while any routed provider/model target uses the singleton estimator. `compactRegion(start, end, agent, signal?)` uses `agent.session` as its single session identity and keeps an optional signal for manual callers. The default summarizer resolves its target from explicit config, the latest logged routed target, then agent options, and records the provider/model pair after any `llm/stream` routing. It replays the routed request's prefix and appends the compaction directive as a trailing user message so the provider's warm KV cache is reused — see the [summary prefix-cache Agent Note](../bug-fix/2026-07-21-compaction-summary-prefix-cache-reuse.md). The call sets the provider-neutral `GenerateOptions.purpose` to `compaction`; adapters may map that purpose to model-hidden transport metadata, and the DeepSeek adapter sends `x-deepseek-harness-compact: 1`.
+`compactIfNeeded(agent, trigger, signal)` takes an explicit `'pressure' | 'context-overflow'` trigger and cancellation. It reads only the latest durable routed request; no header means no work, while any routed provider/model target uses the singleton estimator. `compactNow(agent, signal)` reserves idle turn admission and performs one useful balanced reduction even below pressure, returning `null` without writes when none exists. `compactRegion(start, end, agent, signal?)` uses `agent.session` as its single session identity and keeps an optional signal for explicit callers. The default summarizer resolves its target from explicit config, the latest logged routed target, then agent options, and records the provider/model pair after any `llm/stream` routing. It replays the routed request's prefix and appends the compaction directive as a trailing user message so the provider's warm KV cache is reused — see the [summary prefix-cache Agent Note](../bug-fix/2026-07-21-compaction-summary-prefix-cache-reuse.md). The call sets the provider-neutral `GenerateOptions.purpose` to `compaction`; adapters may map that purpose to model-hidden transport metadata, and the DeepSeek adapter sends `x-deepseek-harness-compact: 1`.
 
 ### Automatic pressure runs after successful durable step work
 
@@ -91,42 +91,44 @@ The basic backend wraps the summary as established checkpoint context and tags i
 
 ### Blocking via a log-recorded lock, plus a crash/recoverable failure taxonomy
 
-The `compact/start … compact/end` bracket is justified, in order of what now does the work:
+The `compact/start … compact/end` bracket is justified by two roles:
 
 1. **Crash-detectable orphan + provenance** (primary). Summarization is a slow model call persisted *after* `compact/start`. A crash mid-summarization leaves a `compact/start` with no matching `compact/end` — a detectable orphan. Releasing the lock last (rather than first) converts the crash window from *silent corruption* into that detectable orphan.
-2. **Prevents concurrent compaction.** `compactRegion` refuses to start if the current turn holds an unmatched `compact/start`. (The loop is single-threaded across either awaited automatic seam, so this is also a re-entry tripwire — a thrown "already in progress" signals a real bug.)
+2. **Prevents concurrent compaction.** Every automatic, manual, and explicit-range entry point refuses a live unmatched `compact/start`. The bracket is the single lock; no process-local mutex duplicates it.
 
-The lock excludes another compaction, not unrelated log-only facts. The basic backend snapshots the token meter's surface nodes after `compact/start` and compares them again after asynchronous summarization; any surface mutation rejects before replacement, while a title or other log-only append leaves the selected span valid.
+The lock excludes another compaction, not unrelated facts. Its markers are time points rather than an exclusive container, so idle injected context may appear between a standalone manual start and end. Automatic work requires whole-surface stability inside its turn. Manual work revalidates only the selected positional span, letting append-only context outside it remain visible after replacement.
 
-Two failure paths, both documented:
+The lifecycle boundary makes crash state unambiguous:
 
-- **Crash** (the loop dies mid-summarization): a dangling `compact/start`, no closer. Because `compact/*` are **log-only**, the orphan is **inert** — no summary replacement lands. The derived surface remains the durable surface present at `compact/start`: full history when pruning made no replacement, or the already-pruned history when it did. Generic turn-repair (`interruptedTurnClosers`) closes the turn with a synthetic `turn/end`; the orphan sits *before* that `turn/end`, so the turn-scoped in-progress check never sees it and a crash cannot wedge future compaction.
-- **Recoverable** (summarization throws but the loop survives): the backend appends `compact/end` with its **`error`** field set and lands no summary replacement. Post-step pressure warns and continues from the latest durable surface — full history if no replacement preceded the attempt, or the pruned surface if pruning already landed. Overflow recovery delegates only before any replacement; generation progress from earlier pruning authorizes a retry from that durable surface unless cancellation or disposal wins.
+- **Current lifecycle:** a dangling `compact/start` after the newest `session/end-seed` is the live durable lock and reports busy.
+- **Later lifecycle:** a newer constructor-written `session/end-seed` proves that the older unmatched start is stale, so resume, fork, and adoption do not remain wedged by a dead writer.
+- **Recoverable failure:** once start lands, the backend makes exactly one `compact/end { error }` attempt. Summary or stability failure leaves the conversation surface unchanged while preserving the failed attempt in the log. If the close append fails, the unmatched start remains intentionally blocking.
 
 `compact/end` keeps its `error?` field (mirroring `tool/result`'s self-contained error — one event tells success from failure without correlating a sibling). There is no separate `compact/error` event.
 
-**Core session repair stays compaction-agnostic — deliberately.** `interruptedTurnClosers` is never taught about `compact/*`. Teaching it would force every future `xxx/start … xxx/end` plugin pair to patch a core module — exactly the coupling the capability-seam architecture exists to avoid. Because the log-only orphan is inert, no special repair is needed: generic turn-repair plus the inertness of an un-landed surface mutation is sufficient.
+**Core session repair stays compaction-agnostic — deliberately.** `interruptedTurnClosers` is never taught about `compact/*`. The general `session/end-seed` lifecycle boundary supplies the evidence the compaction owner needs; the compaction invariant and backend interpret it without adding plugin-specific repair to core.
 
 ## Alternatives considered
 
-- **The full algorithm as concrete interface methods** — rejected because it recouples the contract to one retention strategy. Both core methods are abstract; reusable measurement is a separate LLM-family service and `summarize()` is basic's sole hook.
+- **The full algorithm as concrete interface methods** — rejected because it recouples the contract to one retention strategy. All three operations are abstract; reusable measurement is a separate LLM-family service and `summarize()` is basic's sole hook.
 - **Compaction on `agent/request` or provisional `agent/pre-step` inputs** — rejected because neither proves the final durable request and both couple generic lifecycle to compaction-specific envelope data. Post-step replay plus canonical overflow recovery covers both successful and rejected calls.
 - **A `compact` boolean or untyped request metadata map** — rejected because multiple auxiliary call kinds would become mutually exclusive flags, while an open bag would discard compiler-checked vocabulary. One typed `purpose` discriminant extends with additional call kinds without adding another `GenerateOptions` field.
 - **A separate `compact/error` event** — rejected: `compact/end` keeps an `error?` field, mirroring `tool/result`'s self-contained error — one event tells success from failure without correlating a sibling.
-- **Teaching core turn-repair about `compact/*`** — rejected: the log-only orphan is inert, and a core module patched for every future `xxx/start … xxx/end` plugin pair is exactly the coupling the capability-seam architecture exists to avoid.
+- **Teaching core turn-repair about `compact/*`** — rejected: the general end-seed boundary already distinguishes prior-lifecycle history, and patching core for every future `xxx/start … xxx/end` pair is exactly the coupling the capability-seam architecture exists to avoid.
 
 ## Consequences
 
-- **Packages**: `packages/compact/compact` supplies the interface, `compact-basic` supplies the backend, and `compact-tool-result-prune` supplies optional deterministic rewriting. `packages/llm/token-meter` owns replay-aware measurement independently. The consumer tier is deferred.
+- **Packages**: `packages/compact/compact` supplies the interface, `compact-basic` supplies the backend, `compact-tool-result-prune` supplies optional deterministic rewriting, and `command-compact` supplies human `/compact`. `packages/llm/token-meter` owns replay-aware measurement independently.
 - **Automatic seams**: `agent/post-step` (`@mode serial`) handles successful-call pressure and `agent/request-error` (`@mode waterfall`) handles final request failures after the failed step closes. Generic `agent/pre-step` remains a four-argument checkpoint with no compaction-only prompt/prefix payload.
 - **`SessionEventMap`** gains `compact/start` / `compact/summary` / `compact/end` by declaration merging (merge-extensible); `SurfaceEventType` is **not** touched. These are session events, not cordis `Events`, so the event-taxonomy gate needs no entry.
 - **`dsh-compact`** owns `COMPACT_CHECKPOINT_SOURCE`, `isCompactCheckpointSource(source)`, `toolPairingBalancedBefore(session, seq)`, and `toolPairingBalancedAfter(session, seq)`. The marker identifies replacement summaries across backend implementations. The cached surface-edge checks prevent `compactRegion` and `compactIfNeeded` from splitting a tool-call/result pair, validate current membership by seq, answer both edges from one per-cut balance sequence, and reject stale or missing seqs and orphan results.
-- **`dsh-session`** validates positional replacement, complete provenance, and content-only single-node `tool/result` rewrites through its one surface manager. Its invariant companion treats fresh appended tool results as executions that require an open step and pending call; validated replacements remain turn-enclosed rewrites.
-- **Wiring**: `examples/tui-agent/cordis.yml` loads zero-config `dsh-token-meter`, `dsh-compact-tool-result-prune`, then `dsh-compact-basic`; service-wide defaults make the composition usable without repeated numeric policy.
+- **`dsh-session`** validates positional replacement, complete provenance, and content-only single-node `tool/result` rewrites through its one surface manager. Its invariant companion treats fresh appended tool results as executions that require an open step and pending call, while the compaction companion owns numeric-turn versus standalone-null bracket relations.
+- **Wiring**: `examples/tui-agent/cordis.yml` loads zero-config `dsh-token-meter`, `dsh-compact-tool-result-prune`, `dsh-compact-basic`, then `dsh-command-compact`; service-wide defaults make the composition usable without repeated numeric policy.
 
 ## Testing
 
 - **Unit:** Real Loader and invariant plugins cover whole-unit retention, pruning configuration and replay, rich-block ordering, metadata preservation, convergence, both `compact/end` outcomes, open-tail refusal, pruning-only and summarized overflow recovery, generation proof, caps, and original-error preservation.
 - **Loop:** Tests pin post-step after durable tool results and before `step/end`, actual `agent/request` routing, closed failed steps, fresh retry numbering, and complete thrown/in-band overflow → compaction → reconstructed retry composition.
+- **Manual:** Admission, marker ordering, injection retention, live/stale orphan classification, cancellation, close/flush failures, command mapping, and the queued TUI journey are pinned without a model key.
 - **With-key e2e:** A real model and bash session with lowered limits triggers compaction, records a complete `compact/start…end` pair, shrinks the surface, and finishes the task.
 - **Snapshot gap:** Runaway-turn compaction cannot yet replay because the summarization call records no `assistant/chunk` events or `sessionId`; interleaved summarization-call replay remains follow-up work.

.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.zh.md

@@ -18,10 +18,10 @@ Status: implemented
 
 遵循[能力 seam Agent Note（agent 决策记录）](../architecture/2026-06-13-capability-seams.md)，压缩以独立包（package）发布，使契约、算法和（后续的）消费方 surface 各自独立演进：
 
-1. **接口** — `@deepseek-ai/dsh-compact`：抽象 `CompactService`，拥有 `ctx.compact` 键、`CompactionResult` 词汇、`compact/*` 会话事件以及规范的检查点消息来源。它将 `compactIfNeeded()` 和 `compactRegion()` 声明为**抽象方法**——契约说明压缩*做什么*，而非*怎么做*。
+1. **接口** — `@deepseek-ai/dsh-compact`：抽象 `CompactService`，拥有 `ctx.compact` 键、`CompactionResult` 词汇、`compact/*` 会话事件、手动失败分类体系以及规范的检查点消息来源。它将 `compactIfNeeded()`、`compactNow()` 和 `compactRegion()` 声明为**抽象方法**——契约说明压缩*做什么*，而非*怎么做*。
 2. **实现** — `@deepseek-ai/dsh-compact-basic`：具体的 `BasicCompactService`，消费 `ctx.tokenMeter`，并拥有尾→头保留遍历、通过 `ctx.llm.stream()` 生成摘要、surface 替换、锁、步骤后压力处理和规范的上下文溢出恢复。`summarize()` 是其唯一的子类钩子；计价与回放仍归 meter 所有。
 3. **无模型配套服务** — `@deepseek-ai/dsh-compact-tool-result-prune`：一个具体的可选服务，在后端选择摘要范围之前，重写当前过大的 `tool/result` 节点。它不是第二种压缩实现，也不实现 `CompactService`。
-4. **消费方** — 推迟。一个 `/compact` 工具和斜杠命令将 `inject: ['compact']` 并调用契约；它们被有意排除在本 Agent Note 范围之外，以便 seam 先稳定下来。
+4. **面向用户的消费方** — `@deepseek-ai/dsh-command-compact` 通过 `ctx.commands` 注册无参数 `/compact`，并调用后端无关的 `compactNow()` 操作。它是供用户直接控制的命令，不是面向模型的工具。
 
 ### 契约依赖 `dsh-session` 和 `dsh-llm`——有意为之的偏离
 
@@ -29,11 +29,11 @@ Status: implemented
 
 这不是耦合异味，而是契约的领域所在。「仅 cordis」的指导原则一直是「接口仅依赖契约真正需要命名的东西，绝不依赖实现」的简写。`dsh-session` 和 `dsh-llm` 本身是接口/词汇包，不是实现；`dsh-compact` 仍然不导入任何后端。seam 的真正不变式——*消费方和实现在抽象服务背后独立演进*——完好无损。
 
-### 抽象 `compactIfNeeded` / `compactRegion`，算法在后端
+### 三个抽象操作，算法在后端
 
-早期草案将完整算法（保留遍历、token 求和、文本提取）作为接口上的具体方法。这会将契约重新耦合到一种策略：想要不同保留策略或事件排序的后端必须与继承来的具体代码对抗。将两个核心方法都设为抽象，把所有*怎么做*的决策放在后端，并让接口保持为*做什么*的声明。token 测量根本不是压缩钩子；单例服务使多个消费方能够共享逐会话的回放折叠。
+早期草案将完整算法（保留遍历、token 求和、文本提取）作为接口上的具体方法。这会将契约重新耦合到一种策略：想要不同保留策略或事件排序的后端必须与继承来的具体代码对抗。将三个操作都设为抽象，把所有*怎么做*的决策放在后端，并让接口保持为*做什么*的声明。token 测量根本不是压缩钩子；单例服务使多个消费方能够共享逐会话的回放折叠。
 
-`compactIfNeeded(agent, trigger, signal)` 接受显式的 `'pressure' | 'context-overflow'` 触发原因与取消信号。它只读取最新的持久化已路由请求；没有 header 就不执行工作，任何已路由的提供方/模型目标都使用单例估算器。`compactRegion(start, end, agent, signal?)` 将 `agent.session` 作为唯一会话身份，并为手动调用方保留可选 signal。默认摘要器依次从显式配置、最新记录的已路由目标和 agent 选项解析目标，并在任何 `llm/stream` 路由后记录提供方/模型对。它回放已路由请求的前缀，并将压缩指令追加为尾部 user 消息，从而复用提供方的热 KV cache；见[摘要前缀缓存 Agent Note](../bug-fix/2026-07-21-compaction-summary-prefix-cache-reuse.md)。该调用将提供方无关的 `GenerateOptions.purpose` 设为 `compaction`；适配器可以将此用途映射为对模型隐藏的传输元数据，DeepSeek 适配器会发送 `x-deepseek-harness-compact: 1`。
+`compactIfNeeded(agent, trigger, signal)` 接受显式的 `'pressure' | 'context-overflow'` 触发原因与取消信号。它只读取最新的持久化已路由请求；没有 header 就不执行工作，任何已路由的提供方/模型目标都使用单例估算器。`compactNow(agent, signal)` 会预留空闲轮次接纳，即使未达到压力也进行一次有效的平衡缩减；不存在这种范围时返回 `null`，且不写入任何内容。`compactRegion(start, end, agent, signal?)` 将 `agent.session` 作为唯一会话身份，并为显式调用方保留可选 signal。默认摘要器依次从显式配置、最新记录的已路由目标和 agent 选项解析目标，并在任何 `llm/stream` 路由后记录提供方/模型对。它回放已路由请求的前缀，并将压缩指令追加为尾部 user 消息，从而复用提供方的热 KV cache；见[摘要前缀缓存 Agent Note](../bug-fix/2026-07-21-compaction-summary-prefix-cache-reuse.md)。该调用将提供方无关的 `GenerateOptions.purpose` 设为 `compaction`；适配器可以将此用途映射为对模型隐藏的传输元数据，DeepSeek 适配器会发送 `x-deepseek-harness-compact: 1`。
 
 ### 成功的持久步骤工作完成后运行自动压力检查
 
@@ -91,42 +91,44 @@ compact/end      → log-only. Releases the lock (carries `error` on a recoverab
 
 ### 通过日志记录的锁实现阻塞，加上崩溃/可恢复失败的分类
 
-`compact/start … compact/end` 括号的存在理由，按当前实际承担的职责排序：
+`compact/start … compact/end` 标记对承担两项职责：
 
 1. **可检测的崩溃孤儿 + 来源追溯**（首要）。摘要生成是一次慢速模型调用，持久化在 `compact/start` *之后*。摘要生成中途崩溃会留下一个没有匹配 `compact/end` 的 `compact/start`——一个可检测的孤儿。最后释放锁（而非最先）将崩溃窗口从*静默损坏*转变为可检测的孤儿。
-2. **防止并发压缩。** 如果当前轮次持有未匹配的 `compact/start`，`compactRegion` 拒绝启动。（循环在任一 awaited 自动 seam 上都是单线程的，因此这也是重入绊线——抛出「already in progress」表示真正的 bug。）
+2. **防止并发压缩。** 每个自动、手动和显式范围入口点都会拒绝活动的未匹配 `compact/start`。该标记对就是唯一的锁；没有进程本地 mutex 重复承担同一职责。
 
-该锁只排除另一项压缩，不排除无关的仅日志事实。基础后端会在 `compact/start` 之后对 token meter 的 surface 节点取快照，并在异步摘要后再次比较；任何 surface 变更都会使替换前的检查失败，而标题或其他仅日志追加不会使已选范围失效。
+该锁只排除另一项压缩，不排除无关事实。其标记是时间点，而不是排他的容器，因此空闲注入的上下文可以出现在独立手动 start 与 end 之间。自动工作要求其轮次内的整个 surface 保持稳定。手动工作只重新验证所选位置 span，使其外部的仅追加上下文在替换后保持可见。
 
-两种失败路径，均有文档记录：
+生命周期边界使崩溃状态含义明确：
 
-- **崩溃**（循环在摘要生成中途死亡）：悬空的 `compact/start`，无关闭事件。由于 `compact/*` 是**仅日志**事件，孤儿是**惰性的**，不会落地摘要替换。派生 surface 保持为 `compact/start` 时已经持久化的 surface：如果修剪未产生替换，就是完整历史；如果已经修剪，就是已修剪历史。通用轮次修复（`interruptedTurnClosers`）用合成的 `turn/end` 关闭轮次；孤儿位于该 `turn/end` *之前*，因此轮次范围内的进行中检查永远看不到它，崩溃不会卡住未来的压缩。
-- **可恢复**（摘要生成抛出异常但循环存活）：后端追加设置了 **`error`** 字段的 `compact/end`，但不落地摘要替换。步骤后压力处理发出警告，并从最新的持久 surface 继续：如果尝试前没有替换，就是完整历史；如果修剪已经落地，就是已修剪 surface。溢出恢复只会在没有任何替换前委托；先前修剪带来的 generation 进展允许从该持久 surface 重试，除非取消或资源释放胜出。
+- **当前生命周期：** 最新 `session/end-seed` 之后悬空的 `compact/start` 是活动的持久锁，并报告 busy。
+- **后续生命周期：** 构造函数写入的较新 `session/end-seed` 证明更早的未匹配 start 已陈旧，因此恢复、fork 和接手不会被已死的写入方持续卡住。
+- **可恢复失败：** start 落地后，后端会恰好尝试一次 `compact/end { error }`。摘要或稳定性失败会保持会话 surface 不变，同时在日志中保留失败尝试。如果追加闭合事件失败，未匹配 start 会继续有意阻塞。
 
 `compact/end` 保留其 `error?` 字段（与 `tool/result` 的自包含错误一致——一个事件即可区分成功与失败，无需关联兄弟事件）。没有单独的 `compact/error` 事件。
 
-**核心会话修复保持对压缩无感知——这是有意为之。** `interruptedTurnClosers` 从不被教导 `compact/*`。如果教导它，每个未来的 `xxx/start … xxx/end` 插件对都必须修补核心模块——这恰好是能力 seam 架构存在的意义所要避免的耦合。由于仅日志的孤儿是惰性的，不需要特殊修复：通用轮次修复加上未落地 surface 变更的惰性就足够了。
+**核心会话修复保持对压缩无感知——这是有意为之。** `interruptedTurnClosers` 从不被教导 `compact/*`。通用 `session/end-seed` 生命周期边界提供压缩所有方所需的证据；压缩不变量与后端负责解释它，无需向核心添加插件专属修复。
 
 ## 曾考虑的替代方案
 
-- **完整算法作为接口的具体方法**——否决，因为它将契约重新耦合到一种保留策略。两个核心方法都是抽象的；可复用测量属于单独的 LLM 系列服务，`summarize()` 是 basic 唯一的钩子。
+- **完整算法作为接口的具体方法**——否决，因为它将契约重新耦合到一种保留策略。三个操作都是抽象的；可复用测量属于单独的 LLM 系列服务，`summarize()` 是 basic 唯一的钩子。
 - **在 `agent/request` 或临时 `agent/pre-step` 输入上执行压缩**——否决，因为两者都无法证明最终的持久请求，而且都会将通用生命周期耦合到压缩专属的信封数据。步骤后回放与规范溢出恢复同时覆盖成功和被拒绝的调用。
 - **`compact` 布尔值或无类型的请求元数据 map**——否决，因为多个辅助调用种类会变成互斥标志，而开放 map 会丢弃由编译器检查的词汇。一个类型化的 `purpose` 判别字段可以扩展其他调用种类，而无需再为 `GenerateOptions` 添加字段。
 - **单独的 `compact/error` 事件**——否决：`compact/end` 保留 `error?` 字段，与 `tool/result` 的自包含错误一致——一个事件即可区分成功与失败，无需关联兄弟事件。
-- **教导核心轮次修复识别 `compact/*`**——否决：仅日志的孤儿是惰性的，为每个未来的 `xxx/start … xxx/end` 插件对修补核心模块恰好是能力 seam 架构存在的意义所要避免的耦合。
+- **教导核心轮次修复识别 `compact/*`**——否决：通用 end-seed 边界已经能够区分先前生命周期的历史；为每个未来的 `xxx/start … xxx/end` 插件对修补核心模块，恰好是能力 seam 架构存在的意义所要避免的耦合。
 
 ## 后果
 
-- **包**：`packages/compact/compact` 提供接口，`compact-basic` 提供后端，`compact-tool-result-prune` 提供可选的确定性重写。`packages/llm/token-meter` 独立拥有回放感知的测量。消费方层推迟。
+- **包**：`packages/compact/compact` 提供接口，`compact-basic` 提供后端，`compact-tool-result-prune` 提供可选的确定性重写，`command-compact` 提供面向用户的 `/compact`。`packages/llm/token-meter` 独立拥有回放感知的测量。
 - **自动 seam**：`agent/post-step`（`@mode serial`）处理成功调用的压力，`agent/request-error`（`@mode waterfall`）处理失败步骤关闭后的最终请求失败。通用 `agent/pre-step` 保持为四参数检查点，不携带压缩专属的提示词/前缀 payload。
 - **`SessionEventMap`** 通过可合并扩展的声明合并获得 `compact/start` / `compact/summary` / `compact/end`；`SurfaceEventType` **未被**触及。这些是会话事件，不是 cordis `Events`，因此事件分类门禁无需新增条目。
 - **`dsh-compact`** 拥有 `COMPACT_CHECKPOINT_SOURCE`、`isCompactCheckpointSource(source)`、`toolPairingBalancedBefore(session, seq)` 与 `toolPairingBalancedAfter(session, seq)`。该标记用于跨后端实现识别替换摘要。带缓存的 surface 边缘检查会防止 `compactRegion` 和 `compactIfNeeded` 拆分工具调用/结果对，按 seq 校验当前成员关系，从每个切割点的一条平衡序列回答两侧边缘，并拒绝陈旧或缺失的 seq 与孤立结果。
-- **`dsh-session`** 通过唯一的 surface 管理器校验位置替换、完整溯源信息和仅内容的单节点 `tool/result` 重写。其不变式配套插件将新追加的工具结果视为执行，要求存在已打开的步骤与待处理调用；已校验的替换仍是位于轮次内的重写。
-- **接线**：`examples/tui-agent/cordis.yml` 依次加载零配置的 `dsh-token-meter`、`dsh-compact-tool-result-prune` 和 `dsh-compact-basic`；服务级默认值使组合无需重复数值策略即可使用。
+- **`dsh-session`** 通过唯一的 surface 管理器校验位置替换、完整溯源信息和仅内容的单节点 `tool/result` 重写。其不变式配套插件将新追加的工具结果视为执行，要求存在已打开的步骤与待处理调用，而压缩配套组件拥有数字轮次归属与独立 `null` 归属标记对之间的关系。
+- **接线**：`examples/tui-agent/cordis.yml` 依次加载零配置的 `dsh-token-meter`、`dsh-compact-tool-result-prune`、`dsh-compact-basic`，然后加载 `dsh-command-compact`；服务级默认值使组合无需重复数值策略即可使用。
 
 ## 测试
 
 - **单元测试：** 使用真实 Loader 和 invariant 插件覆盖完整单元保留、修剪配置与回放、富块顺序、元数据保留、收敛、`compact/end` 的两种结果、开放尾部拒绝、仅修剪与带摘要的溢出恢复、generation 证明、上限和原始错误保留。
 - **循环测试：** 测试固定步骤后处理发生在持久工具结果之后、`step/end` 之前，使用实际 `agent/request` 路由，关闭失败步骤，分配新的重试编号，并覆盖完整的抛出/带内溢出 → 压缩 → 重建重试组合。
+- **手动测试：** 无需模型密钥即可固定接纳、标记顺序、注入保留、活动／陈旧未匹配标记分类、取消、闭合／flush 失败、命令映射以及排队 TUI 流程。
 - **带密钥 e2e：** 真实模型和 bash 会话在降低的限制下触发压缩，记录完整的 `compact/start…end` 对，缩小 surface，并完成任务。
 - **快照缺口：** 失控轮次压缩尚无法回放，因为摘要调用未记录 `assistant/chunk` 事件或 `sessionId`；交错摘要调用的回放仍是后续工作。

.agents/notes/implemented/feature/2026-07-24-model-facing-session-query-tools.i18n.yaml

@@ -1,6 +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
-2026-07-24-model-facing-session-query-tools.md: 2a9a20a8b39dea309e759f4eb6ddcdabe25dd8be
-2026-07-24-model-facing-session-query-tools.zh.md: 7fbe746681b329b3e50ae74608a8b2e5167c5ae9
+#   pnpm run verify-translation-pairing --write .agents/notes/implemented/feature/2026-07-24-model-facing-session-query-tools.md
+2026-07-24-model-facing-session-query-tools.md: 82fb70349a94916af2e99b83fcbdac765aae3dd0
+2026-07-24-model-facing-session-query-tools.zh.md: 3ffc142b2a27c612bb8a3238823f536871e5ea17

.agents/notes/implemented/feature/2026-07-24-model-facing-session-query-tools.md

@@ -36,7 +36,7 @@ Session-level results include the latest folded title when available. Each tool
 
 ## Host composition
 
-The consumer is an opt-in plugin. The shipped ACP and TUI apps mount `ctx.sessionQuery` for non-model consumers, while the shared Web/headless composition mounts neither the query service nor the consumer. No shipped composition mounts `@deepseek-ai/dsh-tool-session-query`, so default model requests gain no query prompt or schemas. A composition that opts in also chooses whether to mount the generic timeout and spill policies; the dedicated ACP snapshot fixture mounts both and uses private local spill storage. Generic tool presentation requires no session-query-specific client plugin.
+The consumer is an opt-in plugin. The shipped TUI, Web, and headless compositions mount both `ctx.sessionQuery` and `@deepseek-ai/dsh-tool-session-query` through their shared base, so their default model requests include the query prompt and five schemas; the automation-only ACP composition mounts neither. These compositions also supply the generic timeout and spill policies. The dedicated ACP snapshot fixture mounts the consumer and both policies explicitly, with private local spill storage. Generic tool presentation requires no session-query-specific client plugin.
 
 ## Alternatives considered
 
@@ -48,7 +48,7 @@ The consumer is an opt-in plugin. The shipped ACP and TUI apps mount `ctx.sessio
 
 ## Verification
 
-Package tests pin argument validation, filter translation, timestamp normalization, exact-workspace authorization, parent-filter preauthorization and oracle resistance, changed-observation rejection, service-diagnostic redaction for ordinary and adversarial unknown values, best-effort cyclic-cause logging, logger-failure containment, missing-identity behavior, hidden-boundary pruning, current-step exclusion, internal provider paging, exclusive search and parallel exact-read classification, count caps, exact-signal forwarding, abort-reason preservation, persistence cleanup quiescence, one-scan bounded batch title enrichment, projection-before-dequeue ordering, queued-work suppression, started-worker quiescence, per-header validation, title fallbacks, representative search/trace/read rendering, generic presentation, and disposable registration. Integration coverage uses the real SQLite FTS provider over live and persisted sessions. Shipped configuration, app composition tests, and assembled ACP request-header snapshots prove that the model-facing consumer remains absent while `ctx.sessionQuery` stays available where mounted. A package-owned Loader smoke and dedicated keyless ACP snapshot explicitly mount the consumer with timeout and spill support, pinning its prompt guidance, schemas, and path-independent exact event-read retention behavior.
+Package tests pin argument validation, filter translation, timestamp normalization, exact-workspace authorization, parent-filter preauthorization and oracle resistance, changed-observation rejection, service-diagnostic redaction for ordinary and adversarial unknown values, best-effort cyclic-cause logging, logger-failure containment, missing-identity behavior, hidden-boundary pruning, current-step exclusion, internal provider paging, exclusive search and parallel exact-read classification, count caps, exact-signal forwarding, abort-reason preservation, persistence cleanup quiescence, one-scan bounded batch title enrichment, projection-before-dequeue ordering, queued-work suppression, started-worker quiescence, per-header validation, title fallbacks, representative search/trace/read rendering, generic presentation, and disposable registration. Integration coverage uses the real SQLite FTS provider over live and persisted sessions. Shipped configuration and the TUI/Web composition tests prove that the model-facing consumer is present on the TUI, Web, and headless surfaces, while assembled ACP request-header snapshots prove that the automation surface omits it by default. A package-owned Loader smoke and dedicated keyless ACP snapshot explicitly mount the consumer with timeout and spill support, pinning its prompt guidance, schemas, and path-independent exact event-read retention behavior.
 
 ## Consequences
 

.agents/notes/implemented/feature/2026-07-24-model-facing-session-query-tools.zh.md

@@ -36,7 +36,7 @@ Status: implemented
 
 ## 宿主组合
 
-该消费方是一个需显式启用的插件。发布的 ACP 与 TUI 应用为非模型消费方挂载 `ctx.sessionQuery`，而 Web/headless 共享组合既不挂载查询服务，也不挂载该消费方。发布的组合均未挂载 `@deepseek-ai/dsh-tool-session-query`，因此默认模型请求中不包含查询提示词或 schema。选择启用该插件的组合还要决定是否挂载通用的超时与 spill 策略；专用的 ACP 快照 fixture（测试前置数据）同时挂载这两项策略，并使用私有的本地 spill 存储。通用工具表现无需会话查询专用客户端插件。
+该消费方是一个需显式启用的插件。已交付的 TUI、Web 与无头组合通过共享 base 同时挂载 `ctx.sessionQuery` 和 `@deepseek-ai/dsh-tool-session-query`，因此其默认模型请求包含查询提示词与五个 schema；仅用于自动化的 ACP 组合两者均不挂载。这些组合还提供通用的超时与 spill 策略。专用的 ACP 快照 fixture（测试前置数据）显式挂载该消费方与这两项策略，并使用私有的本地 spill 存储。通用工具表现无需会话查询专用客户端插件。
 
 ## 考虑过的替代方案
 
@@ -48,7 +48,7 @@ Status: implemented
 
 ## 验证
 
-包级测试固定参数校验、过滤条件转换、时间戳规范化、精确工作区授权、父级过滤预授权与抵御预言机探测、变更观测拒绝、普通值与对抗性未知值的服务诊断脱敏、尽力记录循环 cause、日志失败隔离、身份缺失行为、隐藏边界裁剪、当前步骤排除、内部提供方翻页、搜索独占与精确读取并行分类、数量上限、精确信号传递、中止原因保留、持久化清理静止、单次扫描且并发有界的批量标题扩充、先投影再取出下一个任务的顺序、抑制排队工作、等待已启动 worker 静止、逐会话头校验、标题回退、代表性搜索／追踪／读取渲染、通用表现与可释放注册。集成覆盖使用真实 SQLite FTS 提供方查询实时与持久化会话。发布配置、应用组合测试与组装后的 ACP 请求头快照证明：面向模型的消费方仍未挂载，而 `ctx.sessionQuery` 在已经挂载该服务的组合中保持可用。包自身的 Loader 冒烟测试与专用无密钥 ACP 快照显式挂载该消费方，并配套启用超时与 spill 支持，固定其提示词指引、schema 以及与路径无关的精确事件读取保留行为。
+包级测试固定参数校验、过滤条件转换、时间戳规范化、精确工作区授权、父级过滤预授权与抵御预言机探测、变更观测拒绝、普通值与对抗性未知值的服务诊断脱敏、尽力记录循环 cause、日志失败隔离、身份缺失行为、隐藏边界裁剪、当前步骤排除、内部提供方翻页、搜索独占与精确读取并行分类、数量上限、精确信号传递、中止原因保留、持久化清理静止、单次扫描且并发有界的批量标题扩充、先投影再取出下一个任务的顺序、抑制排队工作、等待已启动 worker 静止、逐会话头校验、标题回退、代表性搜索／追踪／读取渲染、通用表现与可释放注册。集成覆盖使用真实 SQLite FTS 提供方查询实时与持久化会话。发布配置与 TUI/Web 组合测试证明面向模型的消费方存在于 TUI、Web 与无头界面，而组装后的 ACP 请求头快照证明自动化界面默认不包含它。包自身的 Loader 冒烟测试与专用无密钥 ACP 快照显式挂载该消费方，并配套启用超时与 spill 支持，固定其提示词指引、schema 以及与路径无关的精确事件读取保留行为。
 
 ## 后果
 

.agents/notes/implemented/feature/2026-07-27-tmux-location-context.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/feature/2026-07-27-tmux-location-context.md
-2026-07-27-tmux-location-context.md: bac5861f7f55c259de04d153115f164d90c415ad
-2026-07-27-tmux-location-context.zh.md: 03cd722381c45604f7aae8f3d0a9f9fbb8b12bb5
+2026-07-27-tmux-location-context.md: 9f5e931e56565b0bf3ee219567c6913c40ef1d97
+2026-07-27-tmux-location-context.zh.md: e47eb43c5a9c64bab4518d118cb4dfd580de64e9

.agents/notes/implemented/feature/2026-07-27-tmux-location-context.md

@@ -12,7 +12,7 @@ tmux exposes this without a daemon: `$TMUX_PANE` names the process's pane, and `
 
 ## Decision
 
-`@deepseek-ai/dsh-tmux-context` is an opt-in function plugin in `packages/context/tmux-context/`, alongside the other bounded request-context enrichments that define neither a tool nor a service. Shipped examples do not mount it because tmux-location disclosure and its token cost are deployment policy.
+`@deepseek-ai/dsh-tmux-context` is an opt-in function plugin in `packages/context/tmux-context/`, alongside the other bounded request-context enrichments that define neither a tool nor a service. The shipped TUI mounts it because terminal-multiplexer context is specific to that surface; `dsh-agent-spine-demo` and the Web/headless surfaces stay silent.
 
 **Pull on the first step of each turn, not a tmux push.** The plugin prepends an `agent/step` listener and acts only when `step === 1`. A pull model needs no background process, no hook installation in the user's tmux, and no teardown; it re-reads current state each turn so a moved, renamed, or re-laid-out pane is picked up naturally. Gating on the first step makes the reading per-turn: a location is stable within a turn, and re-querying every step would add cost without new information. A pane moved mid-turn is reflected on the next turn, which is the accepted tradeoff for the simpler design.
 
@@ -42,7 +42,7 @@ The published `./invariant` companion registers no runtime check: a reading is a
 
 ## Consequences
 
-An agent booted inside tmux now receives its own session/window/pane location and window layout as durable, source-attributed context, updated per turn when the location changes. Deployments opt in through cordis.yml; the default spine and shipped examples stay silent. Outside a real tmux pane — including a terminal that merely inherited `$TMUX`/`$TMUX_PANE` — or without a `ctx.bash` executor, the plugin is inert with no error, so composing it is safe everywhere. Because the reading is one durable `user/message`, it survives compaction as ordinary history, contributes nothing to system-prompt assembly or request headers, and costs at most one two-line message per changed turn. The pull model adds one `tmux display-message` subprocess (through the sandboxed bash seam) on the first step of each turn that is due. The optional interval floor is checked before the query and so suppresses both; an unchanged location is detected only by comparing the returned state, so it suppresses the injection while still paying for the query.
+An agent booted inside tmux now receives its own session/window/pane location and window layout as durable, source-attributed context, updated per turn when the location changes. The shipped TUI opts in; custom deployments may compose the plugin directly. Outside a real tmux pane — including a terminal that merely inherited `$TMUX`/`$TMUX_PANE` — or without a `ctx.bash` executor, the plugin is inert with no error, so composing it is safe everywhere. Because the reading is one durable `user/message`, it survives compaction as ordinary history, contributes nothing to system-prompt assembly or request headers, and costs at most one two-line message per changed turn. The pull model adds one `tmux display-message` subprocess (through the sandboxed bash seam) on the first step of each turn that is due. The optional interval floor is checked before the query and so suppresses both; an unchanged location is detected only by comparing the returned state, so it suppresses the injection while still paying for the query.
 
 ## Testing
 

.agents/notes/implemented/feature/2026-07-27-tmux-location-context.zh.md

@@ -12,7 +12,7 @@ tmux 无需守护进程即可暴露这些信息：`$TMUX_PANE` 标识进程所
 
 ## 决策
 
-`@deepseek-ai/dsh-tmux-context` 是位于 `packages/context/tmux-context/` 的可选启用型函数插件，与其他既不定义工具也不定义服务的有界请求上下文增强并列。随附示例不挂载它，因为 tmux 位置披露及其 token 成本属于部署策略。
+`@deepseek-ai/dsh-tmux-context` 是位于 `packages/context/tmux-context/` 的可选启用型函数插件，与其他既不定义工具也不定义服务的有界请求上下文增强并列。已交付的 TUI 会挂载它，因为终端复用器上下文是该界面特有的；`dsh-agent-spine-demo` 与 Web／无头界面保持沉默。
 
 **在每轮的第一个 step 拉取，而非 tmux 推送。** 插件前置注册一个 `agent/step` 监听器，仅在 `step === 1` 时动作。拉取模型无需后台进程、无需在用户的 tmux 中安装 hook、也无需清理；它每轮重新读取当前状态，因此被移动、改名或重新布局的 pane 都会被自然感知。以第一个 step 为门槛使读数按轮次生成：位置在一轮内是稳定的，逐步骤重复查询只会增加成本而不带来新信息。轮次中途移动的 pane 会在下一轮反映，这是换取更简单设计所接受的取舍。
 
@@ -42,7 +42,7 @@ window active=<0|1>, pane active=<0|1>, layout <window-layout>
 
 ## 后果
 
-启动于 tmux 内的 agent 现在会以持久、带来源标记的上下文收到自身的 session/window/pane 位置及 window 布局，并在位置变化时按轮次更新。部署方通过 cordis.yml 选择启用；默认 spine 与随附示例保持沉默。在真实 tmux pane 之外——包括仅继承了 `$TMUX`／`$TMUX_PANE` 的终端——或没有 `ctx.bash` 执行器时，插件保持惰性且不报错，因此在任何地方组合它都安全。由于读数是一条持久的 `user/message`，它作为普通历史经受压缩，对系统提示装配与请求头毫无贡献，且每个发生变化的轮次至多花费一条两行消息。拉取模型在每个到期轮次的第一个 step 增加一次 `tmux display-message` 子进程（经沙箱化的 bash seam）。可选的间隔下限在查询之前检查，因此同时抑制查询与注入；而位置未变化只能通过比较查询返回的状态得知，因此它只抑制注入，查询开销仍会付出。
+启动于 tmux 内的 agent 现在会以持久、带来源标记的上下文收到自身的 session/window/pane 位置及 window 布局，并在位置变化时按轮次更新。已交付的 TUI 选择启用；自定义部署可直接组合该插件。在真实 tmux pane 之外——包括仅继承了 `$TMUX`／`$TMUX_PANE` 的终端——或没有 `ctx.bash` 执行器时，插件保持惰性且不报错，因此在任何地方组合它都安全。由于读数是一条持久的 `user/message`，它作为普通历史经受压缩，对系统提示装配与请求头毫无贡献，且每个发生变化的轮次至多花费一条两行消息。拉取模型在每个到期轮次的第一个 step 增加一次 `tmux display-message` 子进程（经沙箱化的 bash seam）。可选的间隔下限在查询之前检查，因此同时抑制查询与注入；而位置未变化只能通过比较查询返回的状态得知，因此它只抑制注入，查询开销仍会付出。
 
 ## 测试
 

.agents/notes/implemented/feature/2026-07-30-compaction-progress-visibility.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/implemented/feature/2026-07-30-compaction-progress-visibility.md
+2026-07-30-compaction-progress-visibility.md: b4d95d4bc645924b96eab6a36ee7b8f36b76a2c6
+2026-07-30-compaction-progress-visibility.zh.md: e444fbdfbf1c865a5bb3b6675a5ff318c5cef9e7

.agents/notes/implemented/feature/2026-07-30-compaction-progress-visibility.md

@@ -0,0 +1,49 @@
+# Agent Note: Live standalone compaction progress in the terminal
+
+Status: implemented
+
+English | [中文](2026-07-30-compaction-progress-visibility.zh.md)
+
+## Problem
+
+A standalone manual compaction runs between turns while the agent remains idle. The TUI's turn-phase indicator therefore kept its plain `>` caret throughout the slow summary operation, and a failed attempt produced no transcript row because no replacement checkpoint landed. The liveness presentation needs to reuse the existing status indicator without introducing a second animated status location.
+
+The durable log can retain an unmatched `compact/start` after a process dies. That orphan is useful recovery evidence, but it is not proof that work is running in the current process; replaying it as progress would leave resumed sessions with a permanent phantom indicator.
+
+## Decision
+
+The TUI treats the live standalone `compact/start { turn: null }` to matching `compact/end` bracket as the source of in-flight compaction presentation. A module-local `compacting` cell records the render-clock start and owns one animation timer. A fixed row above the prompt renders `Context being compacted <elapsed>` from that clock, the existing one-cell running status indicator renders `⊙` through the same fade and throb path as turn-phase glyphs, and the terminal progress bit remains active until the bracket closes.
+
+`runningPhaseGlyph` owns the choice among turn-phase glyphs, `⊙`, and the idle caret. Turn-phase glyphs take precedence because numbered compaction brackets are enclosed by a running turn whose phase already lights the indicator. The fixed row is outside the transcript and owns neither a spinner nor another timer; it collapses when empty. The compaction cell does not change the idle editor border, hint, or steering badge, so prompts remain visibly accepted while standalone compaction reserves turn admission.
+
+The cell is live-only. Mount and transcript replay never scan history for an unmatched start; only a `session/event` notification observed by the mounted TUI can open it. Turn-status transitions preserve the cell, while terminal teardown clears its timer and progress bit.
+
+On `compact/end`, the TUI clears the cell before starting the ordinary glyph fade-out. An end carrying `error` adds `Compaction failed: <error>` as a warning. Successful completion remains represented by the landed replacement's transcript marker, and duration remains derivable from the matching durable start and end timestamps without another settled row.
+
+This decision partially supersedes only the progress-related deferred clauses in the [terminal transcript decision](../bug-fix/2026-07-29-human-transcript-append-origin.md) and [browser transcript decision](../bug-fix/2026-07-30-web-transcript-log-ordered-projection.md): progress does not require marker scale or a replacement-rendering refactor. Both notes remain active and continue to own append-origin transcript projection and landed checkpoint markers. The [queued manual compaction decision](2026-07-30-queued-manual-compaction.md) remains the owner of bracket ordering, locking, and stale-orphan classification.
+
+## Alternatives considered
+
+**Add `progressLabel` to `CommandDefinition` and a second TUI status controller, as explored in PR #669.** Rejected because command metadata is not the compaction lifecycle authority, automatic compaction does not originate from a human command, and two status controllers can disagree about the same indicator.
+
+**Add `compacting` to `TurnPhase`, as explored in PR #669.** Rejected because standalone compaction deliberately has no turn, while numbered compaction already has a visible running-turn phase.
+
+**Add a fifth `TimingBucket`.** Rejected because timing buckets partition an open model step and feed its transcript footer. Standalone compaction has no step transition, and a new bucket would add a meaningless compaction column to every step total.
+
+**Share one timer among running, fading, and compaction states.** Rejected because fade-out owns a self-terminating timer, while live compaction has an independent open/close lifetime. Sharing would restructure the reviewed animation state machine without removing an actual concurrent timer.
+
+**Scan the log for an unmatched `compact/start`.** Rejected because a stale orphan from an earlier process lifecycle is expected durable history. Only the live notification proves current work.
+
+**Use a generic command-running indicator.** Rejected for this behavior because the compaction bracket is the more precise source and also covers non-command paths. A future generic command indicator belongs to the `command/run` / `command/done` lifecycle.
+
+**Add an animated compaction row to the transcript.** Rejected because it creates a second status animation point for the same lifecycle. The existing one-cell indicator owns liveness, while the landed marker and failure warning own settled transcript presentation.
+
+**Print a success notice with duration.** Rejected because the landed replacement already supplies the completion marker. The bracket timestamps preserve duration for a future presentation that justifies another transcript row.
+
+## Consequences
+
+Manual compaction has a named elapsed-time display above the prompt while the agent is idle, failure has a direct warning, and a resumed orphan never looks active. The prompt indicator remains one terminal cell wide, while the row and indicator reuse the existing status animation, semantic palette, and terminal-progress lifecycle.
+
+The live cell and timer are additional process-local state, cleared on both bracket close and TUI teardown. This is intentionally not reconstructible presentation state: durable history supplies the successful marker and timing facts, while current-process observation alone supplies liveness.
+
+The package-level TUI tests pin standalone start, elapsed-time refresh, single-indicator presentation, numbered-start exclusion, fade-out, failure warning, idle-status preservation, running-turn precedence, orphaned resume, and timer disposal. The assembled `queued-manual-compact` terminal scenario observes `Context being compacted 1.0s` and `dsh ⊙` while the real summary boundary is held.

.agents/notes/implemented/feature/2026-07-30-compaction-progress-visibility.zh.md

@@ -0,0 +1,49 @@
+# Agent Note: 终端中的实时独立压缩进度
+
+Status: implemented
+
+[English](2026-07-30-compaction-progress-visibility.md) | 中文
+
+## 问题
+
+独立手动压缩（compaction）在轮次之间运行，此时 agent（智能体）保持空闲。因此，在缓慢的摘要操作期间，TUI 的轮次阶段指示器始终显示普通的 `>` 光标；尝试失败时，由于没有替换检查点落地，也不会产生 transcript（文本记录）行。运行状态显示需要复用现有的状态指示器，不能再设置第二处带动画的状态显示。
+
+进程终止后，持久日志中可能保留未匹配的 `compact/start`。该未匹配标记是有用的恢复证据，但无法证明当前进程中有工作正在运行；若将其回放为进度，恢复后的会话便会永久显示虚假的进度指示。
+
+## 决策
+
+TUI 将实时独立的 `compact/start { turn: null }` 与匹配的 `compact/end` 组成的标记对，作为显示进行中压缩状态的真源。模块局部的 `compacting` 状态记录渲染时钟的起始时间，并独占一个动画定时器。提示词上方的固定行根据该时钟渲染 `Context being compacted <elapsed>`，现有的单格运行状态指示器通过与轮次阶段字形相同的明暗渐变和呼吸律动路径渲染 `⊙`，终端进度标志位会保持活跃，直至标记对闭合。
+
+`runningPhaseGlyph` 负责在轮次阶段字形、`⊙` 和空闲光标之间作出选择。轮次阶段字形的优先级更高，因为带编号的压缩标记对处于运行中的轮次内，该轮次的阶段已经激活指示器。该固定行位于 transcript 之外，不带 spinner，也不另设定时器；内容为空时会折叠。压缩状态不会改变空闲编辑器边框、提示或 steering（中途引导）徽标，因此，在独立压缩预留轮次准入期间，界面仍会明确显示提示词已获接纳。
+
+该状态只反映实时事件。挂载和 transcript 回放绝不会扫描历史以查找未匹配的 start；只有已挂载的 TUI 观察到 `session/event` 通知，才能开启它。轮次状态转换会保留该状态，而终端清理会清除其定时器和进度标志位。
+
+收到 `compact/end` 时，TUI 会先清除状态，再启动普通字形的淡出。携带 `error` 的结束事件会以警告形式添加 `Compaction failed: <error>`。成功完成仍由已落地替换项的 transcript 标记呈现；无需再添加已结算的行，也可从匹配且已持久记录的开始与结束时间戳推导持续时间。
+
+本决策仅部分取代[终端 transcript 决策](../bug-fix/2026-07-29-human-transcript-append-origin.md)和[浏览器 transcript 决策](../bug-fix/2026-07-30-web-transcript-log-ordered-projection.md)中与进度相关的延期条款：进度显示不要求标记携带规模信息，也不要求重构替换项渲染。两份记录均保持活动状态，并继续负责基于追加来源的 transcript 投影和已落地检查点标记。[排队式手动压缩决策](2026-07-30-queued-manual-compaction.md)继续负责标记对顺序、锁定机制和陈旧未匹配标记分类。
+
+## 曾考虑的替代方案
+
+**按照 PR（Pull Request）#669 中探索的方案，为 `CommandDefinition` 添加 `progressLabel` 和第二个 TUI 状态控制器。** 不予采用：命令元数据并非压缩生命周期的权威依据，自动压缩并非由人工命令发起，两个状态控制器也可能对同一个指示器给出不一致状态。
+
+**按照 PR #669 中探索的方案，将 `compacting` 添加到 `TurnPhase`。** 不予采用：独立压缩按设计没有轮次，而带编号的压缩已经具有可见的运行轮次阶段。
+
+**添加第五个 `TimingBucket`。** 不予采用：计时分桶用于划分一个开放模型步骤内的时间，并为其 transcript 页脚提供数据。独立压缩没有步骤转换，新分桶会在每个步骤总计中加入一个没有意义的压缩列。
+
+**让运行、淡出和压缩状态共享一个定时器。** 不予采用：淡出过程独占一个会自行终止的定时器，而实时压缩有独立的开启与闭合生命周期。共享定时器会重构已经评审的动画状态机，却不能消除实际存在的并发定时器。
+
+**扫描日志，查找未匹配的 `compact/start`。** 不予采用：来自先前进程生命周期的陈旧未匹配标记是预期的持久历史。只有实时通知才能证明当前进程正在执行工作。
+
+**使用通用的命令运行指示器。** 本行为不采用该方案，因为压缩标记对是更精确的真源，并且还覆盖非命令路径。未来若实现通用命令指示器，应归属于 `command/run` 和 `command/done` 生命周期。
+
+**在 transcript 中添加动画压缩行。** 不予采用：这会为同一个生命周期设置第二处状态动画显示。现有的单格指示器负责显示运行状态，而已落地标记和失败警告负责已结算的 transcript 呈现。
+
+**打印包含持续时间的成功通知。** 不予采用：已落地的替换项已经提供完成标记。标记对的时间戳保留了持续时间，可供未来能够证明新增 transcript 行合理的展示方式使用。
+
+## 后果
+
+手动压缩在 agent 空闲时会在提示词上方显示带名称的已用时间，失败会直接产生警告，恢复会话时的陈旧未匹配标记绝不会显示为活动状态。提示符指示器保持一个终端字符单元宽，状态行和指示器则复用现有的状态动画、语义调色板和终端进度生命周期。
+
+实时状态及其定时器是额外的进程局部状态，在标记对闭合和 TUI 清理这两种情况下都会清除。按设计，这种显示状态不可重建：持久历史提供成功标记与计时事实，只有当前进程的观察才能提供运行中状态。
+
+包（package）级 TUI 测试固定了以下行为：独立开始事件、已用时间刷新、单指示器呈现、排除带编号的开始事件、淡出、失败警告、保留空闲状态、运行轮次优先级、存在未匹配标记时的恢复，以及定时器释放。组装后的 `queued-manual-compact` 终端场景会在真实摘要边界保持开放期间观察到 `Context being compacted 1.0s` 和 `dsh ⊙`。

.agents/notes/implemented/feature/2026-07-30-queued-manual-compaction.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/implemented/feature/2026-07-30-queued-manual-compaction.md
+2026-07-30-queued-manual-compaction.md: 05676ef824bc62ddbdbd8895a325570e91e4bafd
+2026-07-30-queued-manual-compaction.zh.md: b2e40a42451570887a194c215df07e98aa1bd864

.agents/notes/implemented/feature/2026-07-30-queued-manual-compaction.md

@@ -0,0 +1,110 @@
+# Agent Note: Queued manual compaction with one durable lock
+
+Status: implemented
+
+English | [中文](2026-07-30-queued-manual-compaction.zh.md)
+
+## Problem
+
+Automatic compaction protects the context window, but an interactive user also needs a deterministic way to condense accumulated history before pressure policy fires. Sending `/compact` as prompt text would spend a model turn and let the conversation model reinterpret a direct control action. Implementing it inside one UI would duplicate command discovery, lifecycle logging, cancellation, and backend policy.
+
+The human command arrives between turns and must summarize asynchronously. A prompt accepted during that wait must keep its ordinary identity, FIFO position, and wakeup behavior, but it must not derive a request from history that compaction is about to replace. A status check is insufficient: a waking send schedules the driver's claim as a microtask, leaving a same-tick interval where status still reads idle even though the prompt already has right of way.
+
+Compaction also needs one mutual-exclusion fact shared by manual, pressure, overflow, and explicit-range entry points. A process-local flag alone cannot explain a crash-recovered log, while a summarize-first transaction leaves no durable evidence during the expensive interval. Conversely, treating marker pairs as exclusive containers would forbid valid idle injection even though injection is explicitly non-waking and immediate between turns.
+
+This note extends the [compaction capability seam](2026-06-18-compaction-capability-seam.md), the [session end-seed boundary](../architecture/2026-07-30-session-end-seed-log-boundary.md), and the [removal of synthetic log-only turns](../simplification/2026-07-28-remove-synthetic-log-only-turns.md). The supersession audit found partial overlap only: each remains active and owns its broader decision.
+
+## Decision
+
+### `/compact` is a command over a backend-independent seam
+
+`@deepseek-ai/dsh-command-compact` registers one argument-free human command through `ctx.commands`. It calls the third abstract `CompactService` operation, `compactNow(agent, signal)`, and maps the closed `ManualCompactionError` taxonomy (`busy | changed | summary | commit | persistence`) to direct UI results. `command/run` and `command/done` preserve the command lifecycle without entering model history or consuming a model-loop turn.
+
+The command plugin tracks each real handler promise independently of the command executor's abort-aware wait. Its composite lifecycle effect unregisters `/compact` before asynchronously draining handlers that already started, so root teardown reaches quiescence only after backend close and flush work settles.
+
+The seam's `ManualCompactAgentContext` adds only `reserveTurnAdmission()` to the session and routing facts compaction already needs. Retention, balancing, summarization, marker ordering, replacement, and durability remain backend responsibilities.
+
+### Idle turn admission is synchronously reservable
+
+`Agent.reserveTurnAdmission(): (() => void) | undefined` claims the boundary before the next ordinary turn. It succeeds only when the driver is idle, no reservation exists, and no accepted waking item already owns the next turn, including a wake whose claim is still a pending microtask.
+
+The reservation does not create a second queue. Later sends keep their `InboxItemId`, placement, FIFO order, and wakeup facts. `acceptsNextStep` remains false, so waking next-step input becomes an ordinary queued follow-up rather than steering. Release is idempotent and re-arms the existing driver path. `inject()` is not withheld.
+
+`whenIdle()` treats a reservation as unfinished activity, including when it holds a waking item. Lifecycle teardown still drains the driver's own activity promise rather than awaiting an external operation, so disposal can cancel and unwind without depending on the reservation holder.
+
+### One parameterized transaction owns every bracket
+
+`dsh-compact-basic` has one region transaction parameterized by bracket owner (`number | null`), stability rule (whole surface or selected span), and an optional flush. It performs one ordering:
+
+1. validate the selected positional range and inspect the durable tail;
+2. reject a live unmatched compaction marker;
+3. append `compact/start` synchronously;
+4. prepare and await summarization;
+5. revalidate the required stability;
+6. append `compact/summary` and the replacement `user/message`;
+7. make exactly one `compact/end` attempt;
+8. flush when the manual caller requested durability.
+
+Automatic and explicit-region work use the numeric owner recovered from the open turn and require whole-surface stability. Manual work reserves admission first, selects a useful range before the transaction, and writes nothing when selection returns `null`. Its bracket uses `turn: null`, requires only selected-span stability, and flushes every successfully closed attempt before releasing admission in `finally`.
+
+`compact/start` is therefore the only compaction lock. There is no `WeakSet`, wrapper mutex, locked/unlocked method split, or redundant activity check around the transaction.
+
+### Bracket-first deliberately differs from the surveyed implementations
+
+Codex models manual compaction as a `CompactTask` occupying its active-turn slot while automatic compaction runs inline. Pi uses the existence of a compaction abort controller as its mutex and appends compaction only after success. Claude Code shares one compaction routine between automatic and manual paths but constructs its boundary after summary streaming.
+
+DSH deliberately records `compact/start` before calling the summarizer. A slow or crashed attempt is observable, automatic and manual paths share the same durable lock, and a later writer cannot mistake an in-flight summary for an unlocked session. This is a conscious divergence from summarize-first behavior, not an accidental event-order difference.
+
+### Markers are time points, not an event container
+
+`compact/start` and `compact/end` mean lock acquisition and release. They do not claim exclusive ownership of every event between their seqs. An idle `inject()` may append a `user/message` while a manual summary is pending, so that unrelated event can sit inside the marker interval.
+
+Manual stability checks only the selected span: it must remain present, contiguous, ordered, equally priced, and balanced. Append-only context outside it does not stale the summary. Positional replacement places the checkpoint at the old span's surface position and leaves injected context after it in derived model history, even though the injection's log seq precedes the later summary and replacement events.
+
+Failed `changed` or `summary` attempts leave the conversation surface unchanged, but the log is not unchanged: it contains `compact/start` and `compact/end { error }`. User-facing text states that distinction.
+
+### End-seed distinguishes live and stale orphans
+
+Tail scanning finds the current turn, unmatched compaction start, and newest `session/end-seed` independently. An unmatched start after the newest end-seed is live and blocks every compaction entry point. An unmatched start before a later end-seed belongs to an earlier session lifecycle and is stale, so it does not wedge the resumed or forked session.
+
+The compaction invariant uses the same transition logic during seed replay: `session/end-seed` clears an open historical trace. The boundary need not publish live from the constructor for this case; replay is the load-bearing path.
+
+The client request projection closes an unmatched compaction request as interrupted at the `session/end-seed` time and clears its active index. A later `compact/start` therefore creates an independent request instead of leaving or overwriting a permanently running orphan.
+
+Once a transaction has appended its start, every later failure makes one closing attempt. A failed close leaves the unmatched start deliberately visible and blocking, and no flush is attempted. A closed manual attempt is flushed even when it reports an expected failure. Cancellation retains exact-reason precedence after required close and flush cleanup.
+
+### Reference implementation boundaries
+
+[PR #835](https://github.com/deepseek-harness/deepseek-harness/pull/835) was used as a reference implementation for the command, reservation, tests, and snapshot shape, but was not merged. Its process-local `WeakSet` lock and locked/unlocked method splits were considered and not adopted because the durable bracket is the single reachable lock.
+
+That reference also carried client-side replacement-anchor machinery to preserve transcript placement. The log-ordered transcript projection already consumes compaction from event order and does not consult mutable surface positions, so those anchors were considered and not adopted.
+
+## Alternatives considered
+
+**Check `agent.status` without reserving admission.** Rejected because an accepted waking send can still be waiting on its claim microtask while status reads idle.
+
+**Queue the command itself.** Rejected because `/compact` is direct control, not model input, and a prompt already accepted first must retain right of way rather than being reordered around a second command queue.
+
+**Summarize before appending `compact/start`.** Rejected because the expensive in-flight operation would be invisible and would not participate in the lock shared by automatic compaction.
+
+**Use both a durable marker and a process-local mutex.** Rejected because two authorities can disagree after replay and require wrapper branches for states the bracket already expresses.
+
+**Hold injection with waking prompts.** Rejected because idle injection is non-waking durable context by contract; delaying it would make plugin ordering depend on a UI command.
+
+**Require the marker interval to contain only compaction events.** Rejected because markers represent lock time points. Provenance names the selected and shadowed seqs exactly; exclusivity would add no correctness and would reject valid injection.
+
+**Treat every unmatched marker as permanently busy.** Rejected because a crash-recovered or forked session would remain wedged. `session/end-seed` is the explicit lifecycle evidence that distinguishes stale history from a live process-local attempt.
+
+## Verification
+
+Agent-loop tests cover same-tick right of way, preserved IDs and FIFO lifecycle, waking and quiet queued work, idempotent release, `whenIdle()`, cancellation, and teardown. Compact tests cover standalone and numbered invariant ownership, end-seed replay, live versus stale orphans, re-entrant listeners, selected-span drift, commit and close failures, flush ordering, exact cancellation causes, raw output and usage preservation, and automatic/manual mutual exclusion.
+
+The command package pins registration, Loader composition, argument rejection, exact success/failure text, cancellation, absence from model history, and disposal waiting across separate close and flush boundaries after an abort stops the executor from awaiting the handler. The client runtime projection test pins end-seed interruption followed by an independent completed attempt. The `queued-manual-compact` terminal snapshot drives real keystrokes through the assembled TUI: `/help` discovers the command, a held summary admits a queued prompt and immediate injection, `turn: null` markers and the flush precede the queued prompt turn, command lifecycle stays log-only, and the derived order is checkpoint → injection → queued prompt.
+
+## Consequences
+
+Interactive users can compact useful history without spending a conversation-model turn. A prompt accepted before the command wins; one submitted during the command waits with its original queue identity. Manual compaction consumes session seqs but no turn number.
+
+The log exposes slow, failed, crashed, and successful attempts through the same bracket. A stale pre-boundary orphan no longer wedges a new lifecycle, while a current unmatched start remains a hard busy signal. Marker intervals may contain unrelated events, so consumers use provenance and relative ordering rather than assuming a contiguous compaction-only slice.
+
+The shared transaction keeps one ordering and one lock across every entry point. Failure reporting is precise about whether only the log changed, the surface may have partially changed, or the in-memory commit could not be persisted.

.agents/notes/implemented/feature/2026-07-30-queued-manual-compaction.zh.md

@@ -0,0 +1,110 @@
+# Agent Note: 使用单一持久锁实现排队手动压缩
+
+Status: implemented
+
+[English](2026-07-30-queued-manual-compaction.md) | 中文
+
+## 问题
+
+自动压缩（compaction）可以保护上下文窗口，但交互用户还需要一种确定性方法，在压力策略触发前压缩累积的历史。把 `/compact` 作为提示词文本发送会消耗一个模型轮次，还会让会话模型重新解释一项直接控制操作。在某个 UI 内实现该功能，则会重复命令发现、生命周期日志记录、取消与后端策略。
+
+面向用户的命令在轮次之间到达，并且必须异步生成摘要。在等待期间获接纳的提示词必须保留普通身份、FIFO 位置与唤醒行为，但不得从即将被压缩替换的历史派生请求。仅检查状态并不足够：唤醒发送会把驱动器的认领安排为 microtask，因此在同一 tick 内存在一段间隔，此时状态仍显示 idle，但提示词已经拥有优先权。
+
+手动、压力、溢出和显式范围入口点还需要共享同一项互斥事实。仅使用进程本地标志无法解释一份崩溃恢复后的日志，而先摘要再记录的事务在开销较大的等待期间不会留下持久证据。反过来，把标记对视为排他容器又会禁止有效的空闲注入，尽管注入按定义不会唤醒，并且会在轮次之间立即执行。
+
+本 Agent Note 扩展[压缩能力 seam](2026-06-18-compaction-capability-seam.md)、[会话 end-seed 边界](../architecture/2026-07-30-session-end-seed-log-boundary.md)和[移除纯日志事件的合成轮次](../simplification/2026-07-28-remove-synthetic-log-only-turns.md)。取代关系审计只发现部分重叠：三者均保持活动状态，并拥有各自更广泛的决策。
+
+## 决策
+
+### `/compact` 是基于后端无关 seam 的命令
+
+`@deepseek-ai/dsh-command-compact` 通过 `ctx.commands` 注册一个无参数、面向用户的命令。它调用第三个抽象 `CompactService` 操作 `compactNow(agent, signal)`，并把封闭的 `ManualCompactionError` 分类体系（`busy | changed | summary | commit | persistence`）映射为直接 UI 结果。`command/run` 和 `command/done` 保留命令生命周期，同时不进入模型历史，也不消耗模型循环轮次。
+
+命令插件会独立跟踪每个实际处理器 promise，不依赖命令执行器的中止感知等待。其复合生命周期 effect 先注销 `/compact`，再异步等待所有已开始的处理器结算，因此根级 teardown 只有在后端的闭合与 flush 工作结算后才会完全停稳。
+
+该 seam 的 `ManualCompactAgentContext` 只在压缩已需使用的会话与路由事实之上增加 `reserveTurnAdmission()`。保留、平衡、摘要、标记排序、替换与持久性仍由后端负责。
+
+### 可以同步预留空闲轮次接纳
+
+`Agent.reserveTurnAdmission(): (() => void) | undefined` 会在下一个普通轮次之前认领边界。仅当驱动器空闲、没有既存预留，而且尚无已获接纳的唤醒项拥有下一轮次时，它才会成功；仍在等待 microtask 认领的唤醒项也包括在内。
+
+该预留不会创建第二个队列。之后发送的项保留其 `InboxItemId`、位置、FIFO 顺序与唤醒信息。`acceptsNextStep` 保持 false，因此唤醒的 next-step 输入会成为普通的排队 follow-up，而不是 steering（中途引导）。释放操作可幂等调用，并重新启用既有驱动器路径。`inject()` 不受阻塞。
+
+`whenIdle()` 会把预留视为尚未完成的活动，包括预留持有唤醒项的情况。生命周期 teardown 仍会排空驱动器自身的 activity promise，而不会等待外部操作，因此 dispose（资源释放）可以执行取消并完成退出清理，无需依赖预留持有方。
+
+### 一个参数化事务拥有每一对标记
+
+`dsh-compact-basic` 只有一个区域事务，由标记归属值（`number | null`）、稳定性规则（整个 surface 或所选 span）与可选 flush 参数化。它按同一顺序执行：
+
+1. 验证所选位置范围，并检查持久日志尾部；
+2. 拒绝活动的未匹配压缩标记；
+3. 同步追加 `compact/start`；
+4. 准备并等待摘要；
+5. 重新验证所需稳定性；
+6. 追加 `compact/summary` 与替换用的 `user/message`；
+7. 恰好尝试一次 `compact/end`；
+8. 当手动调用方要求持久性时执行 flush。
+
+自动和显式区域工作使用从开放轮次恢复的数字归属值，并要求整个 surface 保持稳定。手动工作会先预留接纳，在进入事务前选择有效范围；选择结果为 `null` 时不写入任何内容。其标记对使用 `turn: null`，只要求所选 span 保持稳定，并在 `finally` 中释放接纳预留前 flush 每次成功闭合的尝试。
+
+因此，`compact/start` 是唯一的压缩锁。不存在 `WeakSet`、包装层 mutex、locked／unlocked 方法拆分，也不存在事务外部重复的活动状态检查。
+
+### 先记录标记有意不同于调研过的实现
+
+Codex 将手动压缩建模为占用其活动轮次槽位的 `CompactTask`，自动压缩则以内联方式运行。Pi 使用压缩 abort controller 是否存在作为 mutex，并仅在成功后追加压缩。Claude Code 的自动和手动路径共享同一个压缩例程，但会在摘要流结束后才构造边界。
+
+DSH 有意在调用摘要器前记录 `compact/start`。缓慢或崩溃的尝试因此可观察，自动与手动路径共享同一个持久锁，之后的写入方也不会把正在生成的摘要误判为未锁定会话。这是对先摘要行为的主动偏离，而不是偶然的事件顺序差异。
+
+### 标记是时间点，而不是事件容器
+
+`compact/start` 和 `compact/end` 表示获取与释放锁。它们不声称排他拥有二者 seq 之间的每个事件。手动摘要等待期间，空闲的 `inject()` 可以追加 `user/message`，因此该不相关事件可能位于标记区间内。
+
+手动稳定性只检查所选 span：它必须仍然存在、连续、有序、计价相同且保持平衡。其外部的仅追加上下文不会使摘要陈旧。位置替换会把检查点放在旧 span 的 surface 位置，并使注入上下文在派生模型历史中位于其后，即使注入的日志 seq 早于后续摘要和替换事件。
+
+失败的 `changed` 或 `summary` 尝试会保持会话 surface 不变，但日志并非没有变化：其中会包含 `compact/start` 和 `compact/end { error }`。面向用户的文本会明确说明这一区别。
+
+### End-seed 区分活动与陈旧的未匹配标记
+
+尾部扫描会分别查找当前轮次、未匹配的 compaction start 与最新 `session/end-seed`。位于最新 end-seed 之后的未匹配 start 是活动锁，会阻塞每个压缩入口点。位于较新 end-seed 之前的未匹配 start 属于更早的会话生命周期，已经陈旧，因此不会卡住恢复或 fork 后的会话。
+
+压缩不变量在 seed 回放期间使用同一项转换逻辑：`session/end-seed` 会清除开放的历史追踪状态。此场景不要求构造函数实时发布该边界；回放才是承重路径。
+
+客户端请求投影会在 `session/end-seed` 时刻将未匹配的压缩请求以中断状态结束，并清除其活动索引。因此，后续 `compact/start` 会创建一个独立请求，而不是让该遗留的未匹配请求永久保持运行状态或将其覆盖。
+
+事务追加 start 后，每次后续失败都会进行一次闭合尝试。闭合失败会有意留下可见且具有阻塞作用的未匹配 start，并且不尝试 flush。已闭合的手动尝试即使报告预期失败也会 flush。完成必需的闭合与 flush 清理后，取消仍保留原始原因优先级。
+
+### 参考实现边界
+
+[PR #835](https://github.com/deepseek-harness/deepseek-harness/pull/835) 用作命令、预留、测试与快照结构的参考实现，但未被合并。它的进程本地 `WeakSet` 锁与 locked／unlocked 方法拆分经过评估后未被采用，因为持久标记对是唯一可达的锁。
+
+该参考实现还包含客户端侧替换锚点机制，用于保留 transcript（文本记录）位置。按日志顺序排列的 transcript 投影已经从事件顺序消费压缩，并且不会查询可变 surface 位置，因此这些锚点经过评估后未被采用。
+
+## 曾考虑的替代方案
+
+**仅检查 `agent.status`，不预留接纳。** 不予采用，因为已获接纳的唤醒发送可能仍在等待其认领 microtask，而状态仍显示 idle。
+
+**把命令本身加入队列。** 不予采用，因为 `/compact` 是直接控制而非模型输入；先获接纳的提示词必须保留优先权，不能围绕第二个命令队列重新排序。
+
+**在追加 `compact/start` 前生成摘要。** 不予采用，因为开销较大的进行中操作将不可见，也不会参与自动压缩共享的锁。
+
+**同时使用持久标记与进程本地 mutex。** 不予采用，因为两项权威在回放后可能产生分歧，还会要求用包装层分支处理标记对已经表达的状态。
+
+**与唤醒提示词一起阻塞注入。** 不予采用，因为按契约，空闲注入是不会唤醒的持久上下文；延迟注入会使插件排序依赖某个 UI 命令。
+
+**要求标记区间只包含压缩事件。** 不予采用，因为标记表示锁的时间点。溯源信息会精确指明所选 seq 与被遮蔽 seq；排他性不会增加正确性，只会拒绝有效注入。
+
+**把每个未匹配标记都永久视为 busy。** 不予采用，因为崩溃恢复或 fork 后的会话会永久卡住。`session/end-seed` 是区分陈旧历史与当前进程活动尝试的显式生命周期证据。
+
+## 验证
+
+Agent loop 测试覆盖同一 tick 内的优先权、保留 ID 与 FIFO 生命周期、会唤醒和静默的排队工作、幂等释放、`whenIdle()`、取消与 teardown。压缩测试覆盖独立与数字形式的不变量 owner、end-seed 回放、活动与陈旧未匹配标记、listener 重入、所选 span 漂移、commit 与闭合失败、flush 顺序、原始取消原因、raw output 与 usage 保留，以及自动／手动互斥。
+
+命令包固定注册行为、Loader 组合、参数拒绝、精确的成功／失败文本、取消、不进入模型历史的保证，以及处置操作在中止使执行器停止等待处理器后，仍会跨越相互独立的闭合与 flush 边界等待该处理器结算。客户端运行时投影测试固定 end-seed 中断，以及随后一次独立尝试的完成。`queued-manual-compact` 终端快照通过已组装 TUI 驱动真实按键：`/help` 可发现该命令；被暂停的摘要会接纳一个排队提示词和即时注入；`turn: null` 标记与 flush 先于排队提示词轮次；命令生命周期保持纯日志；派生顺序固定为检查点 → 注入 → 排队提示词。
+
+## 后果
+
+交互用户无需消耗会话模型轮次即可压缩有效历史。在命令前获接纳的提示词胜出；命令期间提交的提示词会以原有队列身份等待。手动压缩会消耗会话 seq，但不消耗轮次编号。
+
+日志通过同一对标记暴露缓慢、失败、崩溃与成功的尝试。边界前的陈旧未匹配标记不会再卡住新的生命周期，而当前未匹配 start 仍是严格的 busy 信号。标记区间可以包含不相关事件，因此消费方使用溯源信息与相对顺序，而不假定存在连续且仅含压缩事件的切片。
+
+共享事务让每个入口点保持同一种顺序并使用同一把锁。失败报告会精确区分只有日志发生变化、surface 可能部分改变，以及内存 commit 无法持久化这三种情况。

.agents/notes/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.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/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.md
+2026-07-31-even-out-shipped-tool-rosters.md: 316e5045e559e2da162c53d64989ccecfd18b857
+2026-07-31-even-out-shipped-tool-rosters.zh.md: ed39212dc4877f4df1dc1c6e84142b61a866c548

.agents/notes/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.md

@@ -0,0 +1,69 @@
+# Agent Note: Even out the shipped tool rosters
+
+Status: implemented
+
+English | [中文](2026-07-31-even-out-shipped-tool-rosters.zh.md)
+
+## Problem
+
+The two shipped `dsh` surfaces offered different tools for no recorded reason. Session checkpoints, tool-result pruning, the goal tools, and Ralph were in `tui.cordis.yml`; `tool-todo` and, later, web search were in `web.cordis.yml`. Neither surface offered session search, a string-replacement editor, or a repeat-tool guard, though all three exist as packages and none is surface-specific.
+
+The result was a user-visible difference nobody had decided: the same model, asked the same thing, could set a goal on the terminal but not in the browser, and could search the web in the browser but not on the terminal.
+
+## Decision
+
+The rows that are not surface-specific move into [`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml), and three more join them: `tool-session-query`, `tool-str-replace-editor`, and `repeat-tool-guard`. Web search moves there too; its [deployment decision](2026-07-31-web-default-search.md) owns the security boundary while the shared base owns its surface-neutral mount. Both surfaces now assemble the same roster: twenty-five tools on every host, plus `glob` and `grep` when ripgrep is available.
+
+Two rows stay surface-specific. `tmux-context` is TUI-only because a browser surface has no terminal multiplexer to describe. `session-reference` is TUI-only because it drives the shared session-query index from the launcher's process-local path, and the browser sidebar reconciles that index on its own first search.
+
+**This change adds only.** No row is removed from either surface and no existing row's configuration is edited: the executors, the sandbox composition, the access defaults, `tools.mode`, and the workflow tool are exactly what they were. A reader comparing the two catalogs before and after should find additions and nothing else.
+
+### What stays unmounted, and why
+
+Three capabilities stay out on the evidence their own packages record, and are listed here so "we forgot" and "we decided against" stay distinguishable.
+
+**`dsh-tool-cordis`** lets the model write JavaScript and mount it as a temporary plugin. Its README states the limit: "The sandbox is containment for honest code, not a security boundary — host-realm helpers on the sandbox global are reachable, so mount code can reach Node" ([Known limitations](../../../../packages/cordis/tool-cordis/README.md)). The `node:vm` realm lives inside the harness process while `dsh-sandbox-local` confines only the argv it spawns, so on the Web surface both the sandbox and the approval seam are bypassed rather than enforced.
+
+**`dsh-web-fetch-local`** stays unmounted and `dsh-tool-web` keeps `fetch: false`. SSRF protection is deferred in the implementation ([`policy.ts`](../../../../packages/web/web-fetch-local/src/policy.ts) validates protocol, credentials, and length only) and the package says so: "this provider is an SSRF primitive and **must not be enabled** in a deployment that can reach sensitive internal network targets" ([README](../../../../packages/web/web-fetch-local/README.md)). The model chooses the target, which includes the harness's own gateway on loopback, private ranges, and cloud metadata endpoints.
+
+Withholding it narrows the surface without removing the reach: `bash` is mounted, so `curl` gets the same page, as a live run confirmed. What the absence buys is the removal of an argument-shaped request primitive that needs no shell — and with it the accidental path where a summarization request quietly reaches loopback. A deployment that must contain outbound traffic needs a network-level control.
+
+**The LSP trio** stays out for an operational reason rather than a security one: `command` resolves from `PATH` at plugin load, so a missing language server fails the whole boot rather than one tool. It becomes mountable once absence degrades to a skipped registration.
+
+### MCP is a dependency, not a row
+
+`@deepseek-ai/dsh-mcp-client` becomes a runtime dependency of the CLI without a row in any shipped config. The plugin mounts exactly one server per instance and `command` is required, so a default would have to name a third-party server and spawn it as a child process on every launch — outside `ctx.bash`, and therefore outside the sandbox policy the Web surface composes.
+
+The layer that would make MCP a default is the one this repository does not have yet: a bridge that reads a user's server list and mounts one client per entry, the same shape [`dsh-hooks-claude`](../../../../packages/hooks/hooks-claude/README.md) already has for a Claude Code `hooks.json`. Shipping the dependency means an installed `dsh` can mount servers from `$DSH_HOME/config.yaml` today; the CLI README carries the YAML.
+
+## Testing
+
+[`apps/cli/tests/shipped-composition.e2e.ts`](../../../../apps/cli/tests/shipped-composition.e2e.ts) boots the shipped tree through the real Loader in a pseudo-terminal and reads the tool names out of the `request/header` the session log persisted, so the assertion is the catalog the model was actually sent. Its `--config` overlay, [`composition-keyless-tail.cordis.yml`](../../../../apps/cli/tests/fixtures/composition-keyless-tail.cordis.yml), is test isolation only: a network-free adapter and workspace-local session artifacts.
+
+That tail also inserts [`composition-settled.ts`](../../../../apps/cli/tests/fixtures/composition-settled.ts), which announces settled Loader activation on the terminal stream. The TUI renders as soon as its own fiber starts, so a prompt typed at the banner can reach the loop while tool rows and persistence are still activating and assemble a partial catalog; gating the smoke's first prompt on that marker is what makes the assertion deterministic.
+
+The same smoke pins the TUI's unchanged execution posture from the same artifact: `tool-bash` emits its `sandbox_permissions` escalation pair only when the mounted executor has wider modes to escalate to, so asserting its **absence** fails if a later change quietly sandboxes this surface.
+
+[`apps/web/tests/shipped-composition.e2e.ts`](../../../../apps/web/tests/shipped-composition.e2e.ts) covers the Web surface in the built lane, asserting its catalog, that its access default is untouched, and that `workspace-write`'s writable roots include the temp directories — a trap that makes sandbox tests lie when the workspace sits under `/tmp` ([`roots.ts`](../../../../packages/sandbox/sandbox/src/roots.ts)).
+
+`glob` and `grep` are asserted as an all-or-nothing pair rather than fixed members: `dsh-tool-fs-search` probes `command -v rg` at load and registers neither tool without ripgrep, which is a host dependency.
+
+Beyond the committed tests, both surfaces were driven against a real key from the built `apps/cli/lib/bin.js` under plain Node. Every mounted tool executed successfully, including `ralph` and `web_search`; the model never reached `cordis_*` or `mcp_*`, fell back to `grep` when asked for LSP navigation, and used a background `bash` task when asked for a persistent terminal.
+
+## Alternatives considered
+
+**Duplicate the shared rows into both overlays instead of promoting them.** Rejected on the one-home rule: three of the new rows would exist twice with no reason for the copies to diverge, and the next roster change would have to remember both.
+
+**Sandbox the TUI in the same change.** Rejected as a separate decision that does not belong in a roster change: the TUI mounts unrestricted executors, and replacing them alters what an existing surface does rather than what it offers. That decision needs its own evidence — not least because the TUI has no `approval/request` answerer, so an escalation there fails closed instead of prompting.
+
+**Enable Code Mode.** Its trust posture is bash-equivalent by design and its tool calls pass the same `tools/pre-execute` gate as bash, so it is not the same call as the model-code tools above. Rejected here anyway: `both` changes every model-visible request on both surfaces, and `code` replaces the wire rather than adding to it — either is a presentation decision, not a roster one.
+
+**Mount an MCP server by default.** Rejected because a shipped default would have to name one, and any choice spawns a third-party child process on every user's machine outside the sandbox. The dependency ships instead.
+
+## Consequences
+
+The same model gets the same tools on both surfaces, and the difference that existed for no recorded reason is gone. The tests assert the twenty-five unconditional names exactly and require the ripgrep-dependent pair to be either present together or absent together on both sides, so a later change that alters only one surface fails a check instead of shipping quietly.
+
+`apps/cli` gains five workspace dependencies: four the shipped tree now mounts, plus `dsh-mcp-client`, which it does not mount and which exists so an installed `dsh` can.
+
+Nothing about execution changed. The TUI still runs the model's commands through unrestricted executors with no approval seam, and the Web surface still defaults to `danger-full-access`. Both are pinned by assertions in this change, which makes them visible rather than fixed — the sandbox decision is still open.

.agents/notes/implemented/feature/2026-07-31-even-out-shipped-tool-rosters.zh.md

@@ -0,0 +1,69 @@
+# Agent Note: 拉平交付的工具清单
+
+Status: implemented
+
+[English](2026-07-31-even-out-shipped-tool-rosters.md) | 中文
+
+## 问题
+
+两个交付的 `dsh` surface 提供着不同的工具,而没有任何记录说明为什么。会话检查点、工具结果裁剪、goal 工具和 Ralph 在 `tui.cordis.yml`;`tool-todo` 以及后来的 web 搜索在 `web.cordis.yml`。两个 surface 都没有会话搜索、字符串替换编辑器和重复工具守卫,尽管这三者都已成包存在,且没有一个是 surface 专属的。
+
+结果是一处没人做过决定的用户可见差异:同一个模型、同一个请求,在终端上能定目标而在浏览器里不能,在浏览器里能搜网页而在终端上不能。
+
+## 决策
+
+那些并非 surface 专属的行移入 [`base.cordis.yml`](../../../../apps/cli/config/base.cordis.yml),另有三行加入:`tool-session-query`、`tool-str-replace-editor` 和 `repeat-tool-guard`。Web 搜索也一并移入；其[部署决策](2026-07-31-web-default-search.md)负责安全边界，共享 base 则负责与 surface 无关的挂载。两个 surface 现在组装同一份清单：每台宿主上都有二十五个工具，ripgrep 可用时再加上 `glob` 和 `grep`。
+
+有两行仍是 surface 专属。`tmux-context` 只在 TUI,因为浏览器 surface 没有终端复用器可描述。`session-reference` 只在 TUI,因为它以 launcher 的进程本地路径驱动共享的 session-query 索引,而浏览器侧边栏会在自己的首次搜索里重建该索引。
+
+**本次改动只做加法。** 两个 surface 都没有任何一行被移除,也没有任何既有行的配置被编辑:执行器、沙箱组合、访问默认值、`tools.mode` 以及 workflow 工具,全都保持原样。对比改动前后的两份目录,读者应当只看到新增,别无其他。
+
+### 什么保持不挂,以及为什么
+
+有三项能力基于其自身包所记录的证据保持在外,列在这里是为了让「我们忘了」和「我们决定不要」保持可区分。
+
+**`dsh-tool-cordis`** 让模型写一段 JavaScript 并挂成临时插件。它的 README 写明了这个界限:「The sandbox is containment for honest code, not a security boundary — host-realm helpers on the sandbox global are reachable, so mount code can reach Node」([Known limitations](../../../../packages/cordis/tool-cordis/README.md))。`node:vm` 的 realm 就在 harness 进程内,而 `dsh-sandbox-local` 只约束它 spawn 出去的 argv,因此在 Web surface 上,沙箱与批准接缝是被绕过而非被执行。
+
+**`dsh-web-fetch-local`** 保持不挂,`dsh-tool-web` 保持 `fetch: false`。SSRF 防护在实现中是 deferred 状态([`policy.ts`](../../../../packages/web/web-fetch-local/src/policy.ts) 只校验协议、凭据与长度),包里也直说了:「this provider is an SSRF primitive and **must not be enabled** in a deployment that can reach sensitive internal network targets」([README](../../../../packages/web/web-fetch-local/README.md))。目标由模型选择,其中包括 harness 自己跑在环回地址上的网关、内网段和云元数据端点。
+
+弃用它收窄的是接触面而非可达性:`bash` 是挂着的,`curl` 照样能拿到同一个页面——一次真实运行确认了这点。这个缺席买到的是去掉一个无需 shell、以参数成形的请求原语,以及随之而来的那条意外路径:一次「帮我总结这个页面」悄悄打到环回地址。真要收住出站流量的部署需要的是网络层管控。
+
+**LSP 三件套**留在外面是运维原因而非安全原因:`command` 在插件加载时从 `PATH` 解析,因此缺少语言服务器会让整次启动失败,而不只是失去一个工具。等到「缺失」退化为「跳过注册」之后,它就可以挂了。
+
+### MCP 是依赖,不是配置行
+
+`@deepseek-ai/dsh-mcp-client` 成为本 CLI 的运行时依赖,但在任何交付配置里都没有对应的行。该插件每个实例只挂载一台服务器,且 `command` 是必填,因此一个默认值必须点名一台第三方服务器,并在每次启动时把它作为子进程 spawn——不经 `ctx.bash`,因而也在 Web surface 所组合的沙箱策略之外。
+
+真正能让 MCP 成为默认的那一层,恰恰是本仓库尚未拥有的:一个读取用户服务器清单、按条目逐台挂载客户端的桥接,形态与 [`dsh-hooks-claude`](../../../../packages/hooks/hooks-claude/README.md) 读取 Claude Code 的 `hooks.json` 完全相同。交付这个依赖意味着已安装的 `dsh` 今天就能从 `$DSH_HOME/config.yaml` 挂载服务器;CLI README 里给了那段 YAML。
+
+## 测试
+
+[`apps/cli/tests/shipped-composition.e2e.ts`](../../../../apps/cli/tests/shipped-composition.e2e.ts) 在伪终端中通过真实 Loader 启动交付树,并从会话日志持久化的 `request/header` 中读出工具名,因此断言的正是模型实际收到的目录。它传入的 `--config` overlay [`composition-keyless-tail.cordis.yml`](../../../../apps/cli/tests/fixtures/composition-keyless-tail.cordis.yml) 只做测试隔离:一个无网络适配器,以及落在工作区内的会话产物。
+
+该尾部还插入了 [`composition-settled.ts`](../../../../apps/cli/tests/fixtures/composition-settled.ts),它在终端字节流上宣告 Loader 激活已 settle。TUI 在自己的 fiber 一启动就渲染,因此在 banner 处敲下的提示词可能在工具行与持久化仍在激活时就抵达循环,从而组装出不完整的目录;把冒烟的首个提示词 gate 在该标记上,正是断言得以确定的原因。
+
+同一份冒烟还从同一份产物上钉住 TUI 未改变的执行姿态:`tool-bash` 只在挂载的执行器确实有更宽模式可升级时才发出 `sandbox_permissions` 升级参数对,因此断言它的**缺席**会在日后有人悄悄给这个 surface 加上沙箱时失败。
+
+[`apps/web/tests/shipped-composition.e2e.ts`](../../../../apps/web/tests/shipped-composition.e2e.ts) 在构建产物 lane 中覆盖 Web surface,断言它的工具目录、它的访问默认值未被触碰,以及 `workspace-write` 的可写根包含临时目录——一个会让沙箱测试说谎的陷阱,当工作区落在 `/tmp` 下时([`roots.ts`](../../../../packages/sandbox/sandbox/src/roots.ts))。
+
+`glob` 与 `grep` 被作为全有或全无的一对断言,而不是固定成员:`dsh-tool-fs-search` 在加载时探测 `command -v rg`,没有 ripgrep 就两个工具都不注册,这是宿主依赖。
+
+除入库测试外,两个 surface 都以 plain Node 从构建产物 `apps/cli/lib/bin.js` 出发、用真实密钥驱动过。每一个已挂载的工具都执行成功,包括 `ralph` 与 `web_search`;模型从未触达 `cordis_*` 或 `mcp_*`,被要求做 LSP 跳转时退化到 `grep`,被要求开持久终端时用了后台 `bash` 任务。
+
+## 曾考虑的替代方案
+
+**把共享的行复制进两份 overlay,而不是提升到 base。** 基于「一处归属」原则否决:新增行里有三行会存在两份,而这些副本没有任何理由发生分歧,下一次改工具清单还得记着改两处。
+
+**在同一次改动里给 TUI 加沙箱。** 不予采纳，因为这是一个不属于工具清单改动的独立决定：TUI 挂的是不受限执行器，替换它们会改变一个既有 surface 做什么，而非它提供什么。这个决定需要自己的证据——尤其因为 TUI 没有 `approval/request` 的应答方，升级请求在那里会 fail-closed，而不是弹出提示。
+
+**开启 Code Mode。** 它的信任立场按设计与 bash 同级,工具调用要过与 bash 相同的 `tools/pre-execute` 闸门,所以它与上面那些模型写码工具不是同一个判断。在这里仍被否决:`both` 会改变两个 surface 上每一个模型可见请求,而 `code` 是把线路替换而非加一个——两者都是呈现方式的决定,不是工具清单的决定。
+
+**默认挂一台 MCP 服务器。** 否决,因为交付默认值必须点名一台,而任何选择都会在每个用户的机器上、在沙箱之外 spawn 一个第三方子进程。改为交付依赖。
+
+## 后果
+
+同一个模型在两个 surface 上拿到同样的工具,那处没有记录理由的差异消失了。测试会精确断言二十五个无条件提供的名称，并要求依赖 ripgrep 的一对工具在两侧要么同时存在、要么同时缺席，因此日后只改一个 surface 都会让检查失败而不是悄悄发出去。
+
+`apps/cli` 增加五个 workspace 依赖:四个是交付树现在挂载的,外加 `dsh-mcp-client`——它并不被挂载,存在的意义是让已安装的 `dsh` 能挂。
+
+执行相关的一切都没有变。TUI 仍以不受限执行器运行模型的命令且没有批准接缝,Web surface 仍默认 `danger-full-access`。两者都由本次改动中的断言钉住,这让它们变得可见而非被修复——沙箱那个决定仍然悬着。

.agents/notes/implemented/feature/2026-07-31-web-default-search.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/feature/2026-07-31-web-default-search.md
-2026-07-31-web-default-search.md: ddc047a963212cb228da67c6c33128877cacf92c
-2026-07-31-web-default-search.zh.md: 05c30b625953ccd54c127a97b646ad7db75f693b
+2026-07-31-web-default-search.md: d9616c27410bb5be9b385a9aaa56c22f6054eeb1
+2026-07-31-web-default-search.zh.md: 27cd330427669a78c03b939c737b37f79fd7965a

.agents/notes/implemented/feature/2026-07-31-web-default-search.md

@@ -1,4 +1,4 @@
-# Agent Note: Default Web search in the Web/headless composition
+# Agent Note: Default Web search in shipped compositions
 
 Status: implemented
 
@@ -10,7 +10,7 @@ The harness had a complete Web capability family—provider registry, DeepSeek/E
 
 ## Decision
 
-`apps/cli/config/web.cordis.yml` explicitly mounts `dsh-web` with `searchProvider: deepseek-official`, `dsh-web-search-deepseek`, and `dsh-tool-web` with `fetch: false`. It does not mount `dsh-web-fetch-local` or select a fetch provider. The shared overlay makes only `web_search` a default for browser and headless sessions; the TUI composition remains unchanged. The explicit search provider id keeps selection independent of registration order and leaves personal or `--config` overlays able to replace or disable the rows.
+`apps/cli/config/base.cordis.yml` explicitly mounts `dsh-web` with `searchProvider: deepseek-official`, `dsh-web-search-deepseek`, and `dsh-tool-web` with `fetch: false`. It does not mount `dsh-web-fetch-local` or select a fetch provider. The shared base makes only `web_search` a default for TUI, browser, and headless sessions. The explicit search provider id keeps selection independent of registration order and leaves personal or `--config` overlays able to replace or disable the rows.
 
 DeepSeek search uses the same `DEEPSEEK_API_KEY` credential reference as the official conversation adapter. The provider resolves that reference inside every search through the optional `ctx.credentials` service; only a composition without the seam falls back to the launching process environment, and a non-empty literal `apiKey` remains the programmatic last resort. A stored or rotated Web Models key therefore reaches the next search without restarting or retaining the value on the provider. Because `WebSearchProvider.available()` is synchronous, it treats an installed resolver as locally usable and missing dynamic credentials fail the operation with the provider-specific `WEB_PROVIDER_CREDENTIAL_MISSING` code while the stable tool schema stays registered.
 
@@ -26,10 +26,10 @@ The default mount does not create a Web-specific permission policy. `web_search`
 
 **Freeze `process.env.DEEPSEEK_API_KEY` at provider load.** Rejected because the Web Models page writes through `ctx.credentials`; the product's documented first-run path must make the next operation work without a restart.
 
-**Mount Web tools in `base.cordis.yml`.** Rejected because that would also change the TUI deployment. The browser and headless entries already share `web.cordis.yml`; they gain the capability together while TUI remains an explicit later decision.
+**Keep Web tools in `web.cordis.yml`.** Rejected because it preserves an unexplained tool-roster difference between TUI and Web/headless. The rows are not surface-specific, so `base.cordis.yml` is their one home; the [tool-roster decision](2026-07-31-even-out-shipped-tool-rosters.md) records the shared composition.
 
 **Enable search and fetch together.** Rejected because default `web_fetch` would allow model-selected anonymous outbound HTTP(S) retrieval to arbitrary URLs. Search covers discovery; deployments that accept broader retrieval can opt into `dsh-web-fetch-local` and set `dsh-tool-web`'s `fetch` option to `true` in their overlay.
 
 ## Consequences
 
-Web/headless model requests carry only the `web_search` schema and search-only prompt guidance in native mode; Code Mode exposes the same search capability beneath `run_code`. The prompt tells the model to use returned snippets and never advertises the disabled `web_fetch` tool. Search adds a complete auxiliary model call and may use the native server tool multiple times; its exact secret-free request remains reconstructable from the initiating session log. The default offers search-result snippets and source metadata but no arbitrary page retrieval; deployments that need full-page fetch must opt in. The Web snapshot lane boots the shipped tree, drives a replayed `web_search` call through the real DeepSeek provider against a local Messages fixture, asserts the durable auxiliary request and structured result, and pins the settled browser presentation. The real-composition smoke test pins the absence of `web_fetch`; provider tests pin missing, stored, and rotated credential behavior plus literal and ambient compatibility.
+Native model requests on every shipped surface carry only the `web_search` schema and search-only prompt guidance; Web/headless Code Mode exposes the same search capability beneath `run_code`. The prompt tells the model to use returned snippets and never advertises the disabled `web_fetch` tool. Search adds a complete auxiliary model call and may use the native server tool multiple times; its exact secret-free request remains reconstructable from the initiating session log. The default offers search-result snippets and source metadata but no arbitrary page retrieval; deployments that need full-page fetch must opt in. The Web snapshot lane boots the shipped tree, drives a replayed `web_search` call through the real DeepSeek provider against a local Messages fixture, asserts the durable auxiliary request and structured result, and pins the settled browser presentation. The TUI/Web composition smokes pin the shared `web_search` roster and absence of `web_fetch`; provider tests pin missing, stored, and rotated credential behavior plus literal and ambient compatibility.

.agents/notes/implemented/feature/2026-07-31-web-default-search.zh.md

@@ -1,4 +1,4 @@
-# Agent Note: Web／无头组合中的默认 Web 搜索
+# Agent Note: 已交付组合中的默认 Web 搜索
 
 Status: implemented
 
@@ -10,7 +10,7 @@ Status: implemented
 
 ## 决策
 
-`apps/cli/config/web.cordis.yml` 明确挂载 `dsh-web`，配置 `searchProvider: deepseek-official`，同时挂载 `dsh-web-search-deepseek`，并以 `fetch: false` 挂载 `dsh-tool-web`。它不挂载 `dsh-web-fetch-local`，也不选择抓取提供方。共享覆盖层只将 `web_search` 设为浏览器与无头会话的默认工具；TUI 组合保持不变。显式搜索提供方 id 使选择不受注册顺序影响，同时个人覆盖层或 `--config` 覆盖层仍可替换或禁用这些配置项。
+`apps/cli/config/base.cordis.yml` 明确挂载 `dsh-web`，配置 `searchProvider: deepseek-official`，同时挂载 `dsh-web-search-deepseek`，并以 `fetch: false` 挂载 `dsh-tool-web`。它不挂载 `dsh-web-fetch-local`，也不选择抓取提供方。共享 base 只将 `web_search` 设为 TUI、浏览器与无头会话的默认工具。显式搜索提供方 id 使选择不受注册顺序影响，同时个人覆盖层或 `--config` 覆盖层仍可替换或禁用这些配置项。
 
 DeepSeek 搜索使用与官方会话适配器相同的 `DEEPSEEK_API_KEY` 凭据引用。提供方在每次搜索内部通过可选的 `ctx.credentials` 服务解析该引用；只有未挂载该 seam 的组合才会回退到启动进程的环境变量，非空的 `apiKey` 字面值仍作为程序化配置的最后兜底。因此，由 Web 的 Models 页存储或轮换的密钥无需重启即可用于下一次搜索，提供方也无需保留该值。由于 `WebSearchProvider.available()` 是同步方法，它会将已安装解析器视为本地可用；若动态凭据缺失，操作会以提供方专属错误码 `WEB_PROVIDER_CREDENTIAL_MISSING` 失败，而稳定的工具 schema 仍保持注册。
 
@@ -26,10 +26,10 @@ DeepSeek 搜索使用与官方会话适配器相同的 `DEEPSEEK_API_KEY` 凭据
 
 **在提供方加载时固定读取 `process.env.DEEPSEEK_API_KEY`。** 不予采纳：Web Models 页面通过 `ctx.credentials` 写入密钥；产品文档规定的首次运行路径必须保证下一次操作无需重启即可生效。
 
-**在 `base.cordis.yml` 中挂载 Web 工具。** 不予采纳：这也会改变 TUI 部署。浏览器与无头入口已经共享 `web.cordis.yml`；两者会一同获得该能力，是否为 TUI 启用则仍留作后续显式决策。
+**将 Web 工具保留在 `web.cordis.yml` 中。** 不予采纳：这会保留 TUI 与 Web／无头界面之间无法解释的工具清单差异。这些配置行并非界面特有，因此其唯一归属是 `base.cordis.yml`；[工具清单决策](2026-07-31-even-out-shipped-tool-rosters.md)记录了这一共享组合。
 
 **同时启用搜索和抓取。** 不予采纳：默认启用 `web_fetch` 会允许模型自行选择任意 URL，执行匿名出站 HTTP(S) 抓取。搜索负责发现信息；接受更广泛抓取范围的部署可以在覆盖层中选择启用 `dsh-web-fetch-local`，并将 `dsh-tool-web` 的 `fetch` 选项设为 `true`。
 
 ## 后果
 
-Web／无头模型请求在原生模式下只会携带 `web_search` schema，以及仅用于搜索的提示词指引；Code Mode 通过 `run_code` 公开相同的搜索能力。该提示词要求模型使用返回的 snippet，且绝不会向模型提及已禁用的 `web_fetch` 工具。搜索会增加一次完整的辅助模型调用，并可能多次使用原生服务器工具；发起会话的日志仍可精确重建其不含密钥的请求。默认配置会提供搜索结果 snippet 与来源元数据，但不支持任意页面抓取；需要抓取完整页面的部署必须自行选择启用抓取。Web 快照通道会启动已交付配置树，使用本地 Messages fixture（测试前置数据），经由真实 DeepSeek 提供方驱动一次回放的 `web_search` 调用，断言持久化的辅助请求与结构化结果，并固定最终浏览器呈现。真实组合冒烟测试固定了不提供 `web_fetch` 这一事实；提供方测试固定缺失、已存储及已轮换凭据的行为，以及字面值与环境变量的兼容性。
+每个已交付界面的原生模型请求都只会携带 `web_search` schema，以及仅用于搜索的提示词指引；Web／无头 Code Mode 通过 `run_code` 公开相同的搜索能力。该提示词要求模型使用返回的 snippet，且绝不会向模型提及已禁用的 `web_fetch` 工具。搜索会增加一次完整的辅助模型调用，并可能多次使用原生服务器工具；发起会话的日志仍可精确重建其不含密钥的请求。默认配置会提供搜索结果 snippet 与来源元数据，但不支持任意页面抓取；需要抓取完整页面的部署必须自行选择启用抓取。Web 快照通道会启动已交付配置树，使用本地 Messages fixture（测试前置数据），经由真实 DeepSeek 提供方驱动一次回放的 `web_search` 调用，断言持久化的辅助请求与结构化结果，并固定最终浏览器呈现。TUI/Web 组合冒烟测试固定了共享的 `web_search` 清单及不提供 `web_fetch` 这一事实；提供方测试固定缺失、已存储及已轮换凭据的行为，以及字面值与环境变量的兼容性。

.agents/notes/implemented/simplification/2026-07-28-remove-synthetic-log-only-turns.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/simplification/2026-07-28-remove-synthetic-log-only-turns.md
-2026-07-28-remove-synthetic-log-only-turns.md: af4da00f4fe1d7aebff845cd55053bb5b807c979
-2026-07-28-remove-synthetic-log-only-turns.zh.md: 9d72781d6b7cf396a830790d108f4ff25adc816a
+2026-07-28-remove-synthetic-log-only-turns.md: fc76667924ec839301aad993efd996112c9a6b09
+2026-07-28-remove-synthetic-log-only-turns.zh.md: 7520c33e2219c5fe7ab7d8da6312247e42cc69b0

.agents/notes/implemented/simplification/2026-07-28-remove-synthetic-log-only-turns.md

@@ -18,11 +18,11 @@ The generic seam also duplicated domain policy. Its marker map said which plugin
 
 Core session invariants continue to enforce core-owned execution relations: turn and step numbering, enclosure of steering, assistant, tool, todo, and request-header events, and same-step tool call/result pairing. Core permits merge-extensible events between turns because only their declaring plugin knows whether they are execution-scoped or standalone. Plugin invariant companions remain responsible for their own event relations.
 
-The title service appends `session/title` directly after its existing service, revision, cancellation, and live-session checks. The bundled model helper appends its literal `session/title-llm-request` record before dispatch. Persistence observes both through the eager `session/event` path and drains them at ordinary checkpoints and lifecycle teardown; neither append forces a flush merely because it is between turns. A fallback, auxiliary request record, or accepted provider title may therefore appear after `turn/end` and before the next `turn/start`.
+The title service appends `session/title` directly after its existing service, revision, cancellation, and live-session checks. The bundled model helper appends its literal `session/title-llm-request` record before dispatch. Persistence observes both through the eager `session/event` path and drains them at ordinary checkpoints and lifecycle teardown; neither append forces a flush merely because it is between turns. A fallback, auxiliary request record, or accepted provider title may therefore appear after `turn/end` and before the next `turn/start`. Manual compaction uses the same between-turn capability for a `compact/* { turn: null }` bracket, but explicitly flushes the closed attempt because `/compact` promises durability before releasing queued prompt admission.
 
 A session fork may end at any stable event position outside an open turn, not only at `turn/end`. This preserves standalone title and context records in a default fork while still rejecting a prefix cut through active execution.
 
-The historical [universal turn-enclosure decision](../../archived/architecture/2026-06-15-turn-enclosure-invariant.md) remains useful only as the reason the synthetic mechanism was introduced. The [context-injection decision](../architecture/2026-07-24-separate-context-injection-from-turn-execution.md) established the current meaning: one turn represents one model-loop execution.
+The historical [universal turn-enclosure decision](../../archived/architecture/2026-06-15-turn-enclosure-invariant.md) remains useful only as the reason the synthetic mechanism was introduced. The [context-injection decision](../architecture/2026-07-24-separate-context-injection-from-turn-execution.md) established the current meaning: one turn represents one model-loop execution. The [queued manual compaction decision](../feature/2026-07-30-queued-manual-compaction.md) applies that rule to a durable multi-event bracket and owns its marker and admission semantics.
 
 ## Alternatives considered
 
@@ -36,8 +36,8 @@ The historical [universal turn-enclosure decision](../../archived/architecture/2
 
 ## Verification
 
-Core invariant tests accept an unknown plugin event between turns while continuing to reject built-in execution events there. Hook, compaction, plan-mode, Code Mode dispatch, and approval invariant companions replay existing logs and reject the same execution-scoped events before commit when no turn is open. Session-title service tests pin one direct fallback event under concurrent refresh, detached-session rejection, and newest-revision acceptance. JSONL and SQLite round trips preserve a title appended after `turn/end` through the persistence lifecycle drain, and fork tests retain a standalone log-only tail while rejecting boundaries inside an open turn. A keyless assembled ACP snapshot delays the model-backed title until after `turn/end` and pins one standalone provider title with no synthetic turn. Generated API and type-equivalence catalogs contain no removed symbol.
+Core invariant tests accept an unknown plugin event between turns while continuing to reject built-in execution events there. Hook, plan-mode, Code Mode dispatch, and approval invariant companions reject their execution-scoped events when no turn is open; the compaction companion separately accepts a balanced `turn: null` manual bracket between turns and requires numeric owners to match an open turn. Session-title service tests pin one direct fallback event under concurrent refresh, detached-session rejection, and newest-revision acceptance. JSONL and SQLite round trips preserve a title appended after `turn/end` through the persistence lifecycle drain, and fork tests retain a standalone log-only tail while rejecting boundaries inside an open turn. A keyless assembled ACP snapshot delays the model-backed title until after `turn/end` and pins one standalone provider title with no synthetic turn. Generated API and type-equivalence catalogs contain no removed symbol.
 
 ## Consequences
 
-Turn counts and outcomes again describe model-loop executions only. Standalone events consume session seqs, start eager persistence like every other append, and require owners to request an explicit durability barrier only when their operation promises one. Generic plugin mistakes no longer fail under a core default enclosure rule, so each plugin that needs an execution relation must state and test that relation itself. The title capability keeps revision ordering and lifecycle persistence with less core state, no duplicate type seam, and no turn-number collision.
+Turn counts and outcomes again describe model-loop executions only. Standalone events and manual compaction brackets consume session seqs without consuming a turn number, start eager persistence like every other append, and require owners to request an explicit durability barrier only when their operation promises one. Generic plugin mistakes no longer fail under a core default enclosure rule, so each plugin that needs an execution relation must state and test that relation itself. The title capability keeps revision ordering and lifecycle persistence with less core state, and manual compaction gains durable control with no synthetic-turn or turn-number collision.

.agents/notes/implemented/simplification/2026-07-28-remove-synthetic-log-only-turns.zh.md

@@ -18,11 +18,11 @@ Status: implemented
 
 核心会话不变量继续强制核心所属的执行关系：轮次与步骤编号、steering、助手、工具、待办和请求头事件的封闭，以及同一步骤内的工具调用／结果配对。核心允许可合并扩展事件位于轮次之间，因为只有声明它们的插件知道这些事件受执行作用域约束，还是可以独立存在。插件的不变量配套组件仍负责其自身的事件关系。
 
-标题服务会在完成既有的服务状态、修订、取消和实时会话检查后，直接追加 `session/title`。随附模型辅助函数会在发起调用前追加其字面量 `session/title-llm-request` 记录。持久化通过尽快处理的 `session/event` 路径观察两者，并在常规检查点与生命周期 teardown 时排空；二者都不会仅因为位于轮次之间就强制 flush。因此，回退标题、辅助请求记录或已接受的提供方标题可以出现在 `turn/end` 之后、下一个 `turn/start` 之前。
+标题服务会在完成既有的服务状态、修订、取消和实时会话检查后，直接追加 `session/title`。随附模型辅助函数会在发起调用前追加其字面量 `session/title-llm-request` 记录。持久化通过尽快处理的 `session/event` 路径观察两者，并在常规检查点与生命周期 teardown 时排空；二者都不会仅因为位于轮次之间就强制 flush。因此，回退标题、辅助请求记录或已接受的提供方标题可以出现在 `turn/end` 之后、下一个 `turn/start` 之前。手动压缩（compaction）利用同一项轮次间能力记录 `compact/* { turn: null }` 标记对，但会显式 flush 已闭合的尝试，因为 `/compact` 承诺在释放排队提示词接纳预留前完成持久化。
 
 会话 fork 可以结束于开放轮次之外的任意稳定事件位置，而不限于 `turn/end`。这样，默认 fork 会保留独立标题和上下文记录，同时仍拒绝在活跃执行过程中截断前缀。
 
-历史上的[通用轮次封闭决策](../../archived/architecture/2026-06-15-turn-enclosure-invariant.md)如今只适合用于解释为何曾引入合成机制。[上下文注入决策](../architecture/2026-07-24-separate-context-injection-from-turn-execution.md)确立了当前语义：一个轮次表示一次模型循环执行。
+历史上的[通用轮次封闭决策](../../archived/architecture/2026-06-15-turn-enclosure-invariant.md)如今只适合用于解释为何曾引入合成机制。[上下文注入决策](../architecture/2026-07-24-separate-context-injection-from-turn-execution.md)确立了当前语义：一个轮次表示一次模型循环执行。[排队手动压缩决策](../feature/2026-07-30-queued-manual-compaction.md)将该规则应用于持久多事件标记对，并拥有其标记与接纳语义。
 
 ## 曾考虑的替代方案
 
@@ -36,8 +36,8 @@ Status: implemented
 
 ## 验证
 
-核心不变量测试会接受轮次之间的未知插件事件，同时继续拒绝位于该处的内置执行事件。钩子、压缩（compaction）、plan-mode、Code Mode 分发和审批的不变量配套组件会回放既有日志，并在没有开放轮次时，于提交前拒绝相同的执行作用域事件。会话标题服务测试会在并发刷新、会话脱离拒绝和最新修订接受场景下，固定一个直接追加的回退事件。JSONL 和 SQLite 往返测试会通过持久化生命周期排空保留追加在 `turn/end` 之后的标题；fork 测试会保留独立纯日志尾部，同时拒绝位于开放轮次内的边界。一个无密钥、经完整组装的 ACP（Agent Client Protocol）快照会将模型生成的标题延迟到 `turn/end` 之后，并固定一个不含合成轮次的独立提供方标题。生成的 API 和类型等价性目录不含任何已移除符号。
+核心不变量测试会接受轮次之间的未知插件事件，同时继续拒绝位于该处的内置执行事件。钩子、plan-mode、Code Mode 分发和审批的不变量配套组件会在没有开放轮次时拒绝其执行作用域事件；压缩配套组件则另外接受轮次之间平衡的 `turn: null` 手动标记对，并要求数字 owner 匹配一个开放轮次。会话标题服务测试会在并发刷新、会话脱离拒绝和最新修订接受场景下，固定一个直接追加的回退事件。JSONL 和 SQLite 往返测试会通过持久化生命周期排空保留追加在 `turn/end` 之后的标题；fork 测试会保留独立纯日志尾部，同时拒绝位于开放轮次内的边界。一个无密钥、经完整组装的 ACP（Agent Client Protocol）快照会将模型生成的标题延迟到 `turn/end` 之后，并固定一个不含合成轮次的独立提供方标题。生成的 API 和类型等价性目录不含任何已移除符号。
 
 ## 后果
 
-轮次计数和结果重新只描述模型循环执行。独立事件会占用会话 seq，像其他追加一样启动尽快持久化，并且仅当操作承诺持久性时，才要求事件所有方请求显式持久性屏障。通用插件错误不再因核心默认的封闭规则而失败，因此每个需要执行关系的插件都必须自行声明并测试该关系。标题功能保留修订排序和生命周期持久化，同时减少了核心状态，不再重复类型 seam，并消除了轮次编号冲突。
+轮次计数和结果重新只描述模型循环执行。独立事件和手动压缩标记对会占用会话 seq，但不占用轮次编号；它们像其他追加一样启动尽快持久化，并且仅当操作承诺持久性时，才要求事件所有方请求显式持久性屏障。通用插件错误不再因核心默认的封闭规则而失败，因此每个需要执行关系的插件都必须自行声明并测试该关系。标题功能保留修订排序和生命周期持久化，同时减少了核心状态；手动压缩则获得持久控制，不产生合成轮次或轮次编号冲突。

.agents/notes/implemented/testing/2026-07-24-web-gui-browser-e2e-lane.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/testing/2026-07-24-web-gui-browser-e2e-lane.md
-2026-07-24-web-gui-browser-e2e-lane.md: cdb7de52c50733d6650202ee2117916319940738
-2026-07-24-web-gui-browser-e2e-lane.zh.md: b850acf026502d054a9d8b2168f0b4f47f58f39b
+2026-07-24-web-gui-browser-e2e-lane.md: f05fc7268cfb613d0af8240bbb65cb154252a620
+2026-07-24-web-gui-browser-e2e-lane.zh.md: 3fd3805053a570a32e63601d7b039db41365309c

.agents/notes/implemented/testing/2026-07-24-web-gui-browser-e2e-lane.md

@@ -76,13 +76,12 @@ Surveyed AI-chat/agent web UIs and mocking layers (LibreChat, vercel/ai-chatbot
 
 ## Testing
 
-`pnpm run test:web` builds and runs the lane keylessly; `test:web:built` runs it against existing build artifacts. `DSH_SNAPSHOT=record pnpm exec vitest run --config vitest.web.config.ts apps/web/tests/<spec>` records a prompting scenario against the live model, and `DSH_SNAPSHOT=refresh pnpm run test:web` rewrites aria goldens keylessly. CI explicitly selects replay mode. The scaffold hermeticity scenario populates distinct entries in all three ambient skill roots and requires none to enter the assembled catalog. `dsh-llm-replay` unit coverage pins pacing, cancellation, consumption diagnostics, sidecar validation, indexed replacement, and the single append position.
+`pnpm run test:web` builds and runs the lane keylessly; `test:web:built` runs it against existing build artifacts. `DSH_SNAPSHOT=record pnpm exec vitest run --config vitest.web.config.ts apps/web/tests/<spec>` records a prompting scenario against the live model, and `DSH_SNAPSHOT=refresh pnpm run test:web` rewrites aria goldens keylessly. CI explicitly selects replay mode. The live-interactions AUTH scenario pins a non-retryable terminal failure as an inline Chat status carrying the display-safe message and code, verifies that provider-echoed credential fragments stay absent from both Chat and Trajectory, and covers composer recovery plus the `turn/end` error. The scaffold hermeticity scenario populates distinct entries in all three ambient skill roots and requires none to enter the assembled catalog. `dsh-llm-replay` unit coverage pins pacing, cancellation, consumption diagnostics, sidecar validation, indexed replacement, and the single append position.
 
 ## Deferred
 
 - **Web header-class pin**: web fixtures tokenize `{{system}}`/`{{tools}}` everywhere and no scenario pins the web composition's prompt/tool schemas (`TODO(web-header-pin)` — the scaffold `recordFixture` JSDoc marks it). Following the TUI scrub-everywhere precedent; revisit when the web assembly's header diverges from the repl composition it mirrors.
 - **Follow-up-prompt-after-resume scenario**: the history/live stitch path over the real wire; add as its own scenario when that code changes or regresses.
-- **Web error surface**: the client consumes no `agent/error` frames and a pre-chunk failure freezes no partial, so a non-retryable provider failure renders no error copy — the user sees the send simply stop. The AUTH scenario pins the current contract (no crash, composer recovers, turn logged `error`) and `FIXME(web-error-surface)` marks where visible error text gets asserted once the UI grows an error rendering.
 - **Composer steering gesture**: the input locks while running (stop-or-wait), so the steering scenario steers over the wire from the page; `TODO(web-steer-composer)` upgrades the drive step to a real composer gesture when the product grows one.
 - **Drag session reorder**: `workspace.insertSessionBefore` has no browser scenario; it needs two sessions materialized in one workspace plus synthesized HTML5 drag events. Add it when that surface changes or regresses. The inert session Rename/Fork/Delete and workspace Delete menu rows get scenarios when they gain behavior.
 

.agents/notes/implemented/testing/2026-07-24-web-gui-browser-e2e-lane.zh.md

@@ -76,13 +76,12 @@ Web GUI 以一条真实组装链交付——chromium 页面 → client 插件 bu
 
 ## Testing
 
-`pnpm run test:web` 构建并无密钥运行该车道；`test:web:built` 基于现有构建产物运行。`DSH_SNAPSHOT=record pnpm exec vitest run --config vitest.web.config.ts apps/web/tests/<spec>` 对真实模型录制一个发起提示的场景，`DSH_SNAPSHOT=refresh pnpm run test:web` 则无密钥重写 aria 预期输出。CI 显式选择回放模式。scaffold 环境隔离场景会在全部 3 个环境 skill 根目录中分别填入不同条目，并要求这些条目都不得进入组装后的目录。`dsh-llm-replay` 单元覆盖率钉住节奏控制、取消、消费诊断、sidecar 校验、按索引替换与唯一的追加位置。
+`pnpm run test:web` 构建并无密钥运行该车道；`test:web:built` 基于现有构建产物运行。`DSH_SNAPSHOT=record pnpm exec vitest run --config vitest.web.config.ts apps/web/tests/<spec>` 对真实模型录制一个发起提示的场景，`DSH_SNAPSHOT=refresh pnpm run test:web` 则无密钥重写 aria 预期输出。CI 显式选择回放模式。live-interactions AUTH 场景会把不可重试的终态失败钉为 Chat 内联状态，其中携带适合展示的消息与错误码，并验证提供方回显的凭据片段不会出现在 Chat 或 Trajectory 中；该场景同时覆盖输入框恢复与 `turn/end` 错误。scaffold 环境隔离场景会在全部 3 个环境 skill 根目录中分别填入不同条目，并要求这些条目都不得进入组装后的目录。`dsh-llm-replay` 单元覆盖率钉住节奏控制、取消、消费诊断、sidecar 校验、按索引替换与唯一的追加位置。
 
 ## 暂缓
 
 - **Web 头类别钉住**：web fixture 处处 token 化 `{{system}}`/`{{tools}}`，没有场景钉住 web 组合的提示词/工具 schema（`TODO(web-header-pin)`——scaffold 的 `recordFixture` JSDoc 有标记）。沿用 TUI 处处脱敏先例；当 web 组装的请求头与其镜像的 repl 组合进一步分叉时重审。
 - **恢复后追问场景**：真实 wire 上的历史/实时缝合路径；当该代码变更或回归时作为独立场景补充。
-- **Web 错误表面**：客户端不消费任何 `agent/error` 帧，分片前的失败也没有可冻结的部分输出，因此不可重试的提供方失败不渲染任何错误文案——用户看到的只是发送就此停住。AUTH 场景钉住当前契约（不崩溃、输入框恢复可用、轮次记录为 `error`），`FIXME(web-error-surface)` 标记了待 UI 长出错误渲染后断言可见错误文本的位置。
 - **输入框 steering 手势**：输入在运行期间锁定（只能停止或等待），因此 steering 场景从页面走 wire 做 steer；`TODO(web-steer-composer)` 待产品长出真实的输入框手势后，把驱动步骤升级为该手势。
 - **拖拽会话重排**：`workspace.insertSessionBefore` 尚无浏览器场景；它需要在同一个工作区里物化两个会话，并合成 HTML5 拖拽事件。当该表面变更或回归时再补充。无行为的会话 Rename/Fork/Delete 和工作区 Delete 菜单行待获得行为后再补充场景。
 

apps/cli/README.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 apps/cli/README.md
-README.md: aee08225febe34bb10ee1001c798db9d7daa8905
-README.zh.md: 8f38fd342a0433ca5f7645f896b81aa5392033ae
+README.md: 43524950ae7d3379406a0c859b6449f0697d204b
+README.zh.md: 01954f1229a6f4d076155d42566ed29d568631ca

apps/cli/README.md

@@ -12,6 +12,7 @@ The TUI surface:
 - treats the **invoking directory** as the workspace — sessions, relative paths, and workspace instructions resolve from the cwd (`dsh meta` is the sole exception, below);
 - tells the agent where its own source lives: after boot it adds a prompt section naming this harness checkout, resolved from the launcher's real path so it holds under a PATH symlink and an arbitrary cwd, so the self-referential `cordis` toolset can read and modify it;
 - applies the personal overlay from `~/.dsh` (see [app-boot's Personal config](../../packages/ui/app-boot/README.md#personal-config)): `config.yaml` patches the booted tree, while `.env` there is the credential provider's own store (never hoisted into the environment, so keys stay rotatable). Environment precedence is ambient > project `.env`.
+- registers bare `/compact`: while the agent is idle, it summarizes useful older history even below automatic pressure, rejects arguments, and reports success only after the standalone replacement bracket is durable. A prompt submitted during compaction keeps its queue identity and starts after that checkpoint; injected context remains visible.
 
 `dsh meta` is that same TUI with this harness checkout as the workspace, so working on dsh itself needs no `cd`. It chdirs to the checkout root — resolved from the launcher's real path, the same root the source-path prompt section names — after the environment is settled, so precedence is unchanged while the session cwd and HMR watch root move together. Meta always starts a fresh session and accepts no default-surface options; use ordinary `dsh --resume <id>` to resume a persisted session.
 
@@ -23,9 +24,9 @@ The Web and headless surfaces boot `base.cordis.yml` plus `web.cordis.yml`, then
 
 The shipped TUI and Web compositions register the native DeepSeek adapter plus pi-ai OpenAI and Anthropic profiles. Credentials and endpoint overrides come from the provider-standard `DEEPSEEK_API_KEY` / `DEEPSEEK_BASE_URL`, `OPENAI_API_KEY` / `OPENAI_BASE_URL`, and `ANTHROPIC_API_KEY` / `ANTHROPIC_BASE_URL` pairs in the boot's layered environment.
 
-The Web/headless composition also registers only `web_search`. Search uses DeepSeek's Anthropic-compatible Messages endpoint, resolves the same `DEEPSEEK_API_KEY` reference for every call, and accepts the separate `DEEPSEEK_SEARCH_BASE_URL` endpoint override; each search is an auxiliary model request with its own latency and token cost. `web_fetch` remains disabled and the composition mounts no default fetch provider, so deployments that need arbitrary page retrieval must opt in through an overlay. The TUI composition does not mount Web tools by default. The deployment decision and its security boundary live in the [default Web search Agent Note](../../.agents/notes/implemented/feature/2026-07-31-web-default-search.md).
+Every surface also registers `web_search` and only `web_search`. Search uses DeepSeek's Anthropic-compatible Messages endpoint, resolves the same `DEEPSEEK_API_KEY` reference for every call, and accepts the separate `DEEPSEEK_SEARCH_BASE_URL` endpoint override; each search is an auxiliary model request with its own latency and token cost. `web_fetch` remains disabled and the composition mounts no default fetch provider, so deployments that need arbitrary page retrieval must opt in through an overlay. The deployment decision and its security boundary live in the [default Web search Agent Note](../../.agents/notes/implemented/feature/2026-07-31-web-default-search.md).
 
-`DSH_TOOLS_MODE` selects the tool presentation mode for the whole Web/headless process: `native` (the schema default when unset), `code` (the `run_code`-only Code Mode wire), or `both`; any other value fails loud at boot through the `dsh-tools` config schema. It is a TEMPORARY seam — process-wide because Loader composition is static — and is removed once the web UI owns per-session tool-mode selection; the TUI surface ignores it (its config tree pins its own mode).
+`DSH_TOOLS_MODE` selects the tool presentation mode for the whole Web/headless process: `native` (the schema default when unset), `code` (the `run_code`-only Code Mode wire), or `both`; any other value fails loud at boot through the `dsh-tools` config schema. It is a TEMPORARY seam — process-wide because Loader composition is static — and is removed once the web UI owns per-session tool-mode selection; the TUI surface ignores it and pins `native`.
 
 [`core-web.cordis.yml`](config/core-web.cordis.yml) is an opt-in `dsh web --config` overlay that keeps the shipped Web host, browser, Workspace, persistence, and permission composition while reducing the default native model surface to owner-scoped persistent `bash` and `str_replace_editor`. The PTY backend and editor consume the existing Web sandbox and filesystem providers. An open persistent shell prevents changing that session's permission mode until the shell closes, so a shell created under wider access cannot survive a downgrade. `DSH_TOOLS_MODE` still controls native/Code Mode presentation for the resulting two-tool registry.
 
@@ -37,6 +38,23 @@ pnpm run dsh web --config apps/cli/config/core-web.cordis.yml
 
 Every `dsh` surface — TUI, Web, and headless — reports session telemetry by default (the row lives in the shared `base.cordis.yml`): every session-log event streams as OTLP/HTTP log records to `https://harness-telemetry.deepseeksvc.com/v1/logs` on a 10-second batch cadence. `DSH_TELEMETRY_OTLP_URL` points the exporter at a different collector; setting `DSH_TELEMETRY_DISABLED` to ANY non-empty value — including `0` or `false` — disables the row before it loads (a privacy switch prefers off-by-mistake over on-by-mistake). No redaction rule is mounted in this composition yet: exported records are the raw captured copy, including message text, tool arguments and results, and the session's working-directory path. The deployment rulings live in the [web-telemetry-default-mount Agent Note](../../.agents/notes/implemented/feature/2026-07-31-web-telemetry-default-mount.md).
 
+MCP servers are not a shipped default, because a default would have to name one: `@deepseek-ai/dsh-mcp-client` mounts exactly one server per row and spawns it as a child process, outside `ctx.bash` and so outside the sandbox policy. The package is a runtime dependency of this CLI, so an installed `dsh` can mount your own servers from `$DSH_HOME/config.yaml` or a `--config` overlay without a source checkout:
+
+```yaml
+- insert:
+    - id: mcp-github
+      name: '@deepseek-ai/dsh-mcp-client'
+      config:
+        serverName: github
+        transport: stdio
+        command: npx
+        args: ['-y', '@modelcontextprotocol/server-github']
+        env:
+          GITHUB_TOKEN: !!js process.env.GITHUB_TOKEN
+```
+
+The model then sees `mcp__github__*`. See the [mcp-client README](../../packages/mcp/mcp-client/README.md) for the Streamable HTTP transport and the full field table.
+
 ## Install (developer machine)
 
 Symlink the source-running launcher onto your PATH; it resolves the checkout through its own real path, so code changes apply on the next launch with no build step:

apps/cli/README.zh.md

@@ -12,6 +12,7 @@ TUI 界面：
 - 将 **调用目录** 视为 workspace：会话、相对路径和 workspace 指令都从 cwd 解析（`dsh meta` 是唯一例外，见下文）；
 - 告知 agent 自身源码所在位置：启动后添加一个命名此 harness checkout 的提示词段。该路径从启动器的真实路径解析，因此在 PATH 符号链接和任意 cwd 下仍然有效，使自指的 `cordis` 工具集可以读取并修改它；
 - 应用 `~/.dsh` 中的个人覆盖（参见 [app-boot 的个人配置](../../packages/ui/app-boot/README.md#personal-config)）：`config.yaml` 修补已启动的树，而那里的 `.env` 是凭据 provider 自己的存储（绝不会被提升进环境，因此密钥始终可轮换）。环境优先级为环境中已有的值 > 项目 `.env`。
+- 注册裸 `/compact`：agent 空闲时，即使未达到自动压力，也会摘要有效的较早历史；该命令拒绝参数，并只在独立替换标记对持久化后报告成功。压缩（compaction）期间提交的提示词保留其队列身份，并在该检查点之后启动；注入的上下文仍保持可见。
 
 `dsh meta` 是以本 harness checkout 为 workspace 的同一个 TUI，因此开发 dsh 自身无需 `cd`。它在环境确定之后才 chdir 到 checkout 根目录（从启动器的真实路径解析，与源码路径提示词段所指的根目录相同），因此环境优先级不变，而会话 cwd 与 HMR 监视根目录会一并移动。Meta 始终创建新会话，不接受默认界面的任何选项；恢复已持久化会话应使用普通的 `dsh --resume <id>`。
 
@@ -23,9 +24,9 @@ Web 和无头界面启动 `base.cordis.yml` 与 `web.cordis.yml`，随后应用
 
 已交付的 TUI 和 Web 组合会注册原生 DeepSeek 适配器，以及 pi-ai 的 OpenAI 和 Anthropic 提供方配置。凭据和端点覆盖来自启动分层环境中的提供方标准变量对：`DEEPSEEK_API_KEY` / `DEEPSEEK_BASE_URL`、`OPENAI_API_KEY` / `OPENAI_BASE_URL` 和 `ANTHROPIC_API_KEY` / `ANTHROPIC_BASE_URL`。
 
-Web／无头组合还只会注册 `web_search`。搜索使用 DeepSeek 的 Anthropic 兼容 Messages 端点，每次调用都会解析同一个 `DEEPSEEK_API_KEY` 凭据引用，并接受独立的 `DEEPSEEK_SEARCH_BASE_URL` 端点覆盖；每次搜索都是一次辅助模型请求，会产生独立的延迟与 token 成本。`web_fetch` 仍处于禁用状态，组合也未挂载默认抓取提供方；需要任意页面抓取能力的部署必须通过覆盖层选择启用。TUI 组合默认不挂载 Web 工具。部署决策及其安全边界见[默认 Web 搜索 Agent Note](../../.agents/notes/implemented/feature/2026-07-31-web-default-search.md)。
+每个界面也都只注册 `web_search` 这一个 Web 工具。搜索使用 DeepSeek 的 Anthropic 兼容 Messages 端点，每次调用都会解析同一个 `DEEPSEEK_API_KEY` 凭据引用，并接受独立的 `DEEPSEEK_SEARCH_BASE_URL` 端点覆盖；每次搜索都是一次辅助模型请求，会产生独立的延迟与 token 成本。`web_fetch` 仍处于禁用状态，组合也未挂载默认抓取提供方；需要任意页面抓取能力的部署必须通过覆盖层选择启用。部署决策及其安全边界见[默认 Web 搜索 Agent Note](../../.agents/notes/implemented/feature/2026-07-31-web-default-search.md)。
 
-`DSH_TOOLS_MODE` 为整个 Web／无头进程选择工具呈现模式：可选值为 `native`（未设置时的 schema 默认值）、`code`（仅含 `run_code` 的 Code Mode 协议接口）或 `both`；任何其他值都会经由 `dsh-tools` 配置 schema 在启动时明确报错。它是一个临时 seam：Loader 组合是静态的，因此该设置作用于整个进程；待 Web UI 负责逐会话工具模式选择后便会移除。TUI 界面会忽略该变量（其配置树固定了自身模式）。
+`DSH_TOOLS_MODE` 为整个 Web／无头进程选择工具呈现模式：`native`（未设置时的 schema 默认值）、`code`（仅含 `run_code` 的 Code Mode 线路）或 `both`；任何其他值都会经由 `dsh-tools` 配置 schema 在启动时明确报错。它是一个临时 seam——Loader 组合是静态的，因此该设置作用于整个进程——待 Web UI 负责逐会话工具模式选择后便会移除；TUI 界面会忽略该变量并固定为 `native`。
 
 [`core-web.cordis.yml`](config/core-web.cordis.yml) 是一个可选启用的 `dsh web --config` 覆盖层：它保留已交付的 Web 宿主、浏览器、Workspace、持久化与权限组合，同时将默认的原生模型界面精简为以所有者为作用域的持久 `bash` 以及 `str_replace_editor`。PTY 后端和编辑器分别消费现有的 Web 沙箱与文件系统提供方。持久 shell 处于打开状态时，会阻止所属会话更改权限模式；因此，在较宽权限下创建的 shell 无法在降权后继续存活。`DSH_TOOLS_MODE` 仍控制由此得到的双工具注册表采用原生／Code Mode 呈现。
 
@@ -37,6 +38,23 @@ pnpm run dsh web --config apps/cli/config/core-web.cordis.yml
 
 每个 `dsh` 界面——TUI、Web 与无头——都默认上报会话遥测（该行位于共享的 `base.cordis.yml`）：每条会话日志事件以 OTLP/HTTP 日志记录的形式、按 10 秒批处理节奏流向 `https://harness-telemetry.deepseeksvc.com/v1/logs`。`DSH_TELEMETRY_OTLP_URL` 可将 exporter 指向其他 collector；将 `DSH_TELEMETRY_DISABLED` 设为**任意非空值**——包括 `0` 或 `false`——都会在该行加载前将其关停（隐私开关取「宁可误关、不可误开」）。该组合当前未挂载任何脱敏规则：导出记录即原始捕获副本，包含消息正文、工具参数与结果、以及会话工作目录路径。部署口径见 [web-telemetry-default-mount Agent Note](../../.agents/notes/implemented/feature/2026-07-31-web-telemetry-default-mount.md)。
 
+MCP 服务器不是交付默认值,因为默认值必须点名一台:`@deepseek-ai/dsh-mcp-client` 每一行只挂载一台服务器,并把它作为子进程 spawn,该进程不经 `ctx.bash`,因此也不受沙箱策略约束。该包是本 CLI 的运行时依赖,所以已安装的 `dsh` 无需源码检出即可从 `$DSH_HOME/config.yaml` 或 `--config` 覆盖层挂载你自己的服务器:
+
+```yaml
+- insert:
+    - id: mcp-github
+      name: '@deepseek-ai/dsh-mcp-client'
+      config:
+        serverName: github
+        transport: stdio
+        command: npx
+        args: ['-y', '@modelcontextprotocol/server-github']
+        env:
+          GITHUB_TOKEN: !!js process.env.GITHUB_TOKEN
+```
+
+模型随后会看到 `mcp__github__*`。Streamable HTTP 传输与完整字段表见 [mcp-client README](../../packages/mcp/mcp-client/README.md)。
+
 ## 安装（开发机）
 
 将从源码运行的启动器符号链接到 PATH 上；它通过自身真实路径解析 checkout，因此代码更改会在下次启动时生效，无需构建：

apps/cli/composition.md

@@ -78,6 +78,8 @@ flowchart LR
   cfg --> plugin_tui_token_meter
   plugin_tui_compact_basic["compact-basic<br/>@deepseek-ai/dsh-compact-basic"]
   cfg --> plugin_tui_compact_basic
+  plugin_tui_command_compact["command-compact<br/>@deepseek-ai/dsh-command-compact"]
+  cfg --> plugin_tui_command_compact
   plugin_tui_subagent["subagent<br/>@deepseek-ai/dsh-subagent"]
   cfg --> plugin_tui_subagent
   plugin_tui_subagent_spawn["subagent-spawn<br/>@deepseek-ai/dsh-subagent-spawn"]
@@ -98,6 +100,28 @@ flowchart LR
   cfg --> plugin_tui_spill_local
   plugin_tui_spill_policy["spill-policy<br/>@deepseek-ai/dsh-spill-policy"]
   cfg --> plugin_tui_spill_policy
+  plugin_tui_session_checkpoint_policy["session-checkpoint-policy<br/>@deepseek-ai/dsh-session-checkpoint-policy"]
+  cfg --> plugin_tui_session_checkpoint_policy
+  plugin_tui_tool_result_prune["tool-result-prune<br/>@deepseek-ai/dsh-compact-tool-result-prune"]
+  cfg --> plugin_tui_tool_result_prune
+  plugin_tui_tool_todo["tool-todo<br/>@deepseek-ai/dsh-tool-todo"]
+  cfg --> plugin_tui_tool_todo
+  plugin_tui_tool_goal["tool-goal<br/>@deepseek-ai/dsh-tool-goal"]
+  cfg --> plugin_tui_tool_goal
+  plugin_tui_tool_ralph["tool-ralph<br/>@deepseek-ai/dsh-tool-ralph"]
+  cfg --> plugin_tui_tool_ralph
+  plugin_tui_tool_session_query["tool-session-query<br/>@deepseek-ai/dsh-tool-session-query"]
+  cfg --> plugin_tui_tool_session_query
+  plugin_tui_tool_str_replace_editor["tool-str-replace-editor<br/>@deepseek-ai/dsh-tool-str-replace-editor"]
+  cfg --> plugin_tui_tool_str_replace_editor
+  plugin_tui_repeat_tool_guard["repeat-tool-guard<br/>@deepseek-ai/dsh-repeat-tool-guard"]
+  cfg --> plugin_tui_repeat_tool_guard
+  plugin_tui_web["web<br/>@deepseek-ai/dsh-web"]
+  cfg --> plugin_tui_web
+  plugin_tui_web_search_deepseek["web-search-deepseek<br/>@deepseek-ai/dsh-web-search-deepseek"]
+  cfg --> plugin_tui_web_search_deepseek
+  plugin_tui_tool_web["tool-web<br/>@deepseek-ai/dsh-tool-web"]
+  cfg --> plugin_tui_tool_web
   plugin_tui_tools["tools<br/>@deepseek-ai/dsh-tools"]
   cfg --> plugin_tui_tools
   plugin_tui_system_prompt["system-prompt<br/>@deepseek-ai/dsh-system-prompt"]
@@ -147,6 +171,7 @@ flowchart LR
 | `plan-mode` | `@deepseek-ai/dsh-plan-mode` |
 | `token-meter` | `@deepseek-ai/dsh-token-meter` |
 | `compact-basic` | `@deepseek-ai/dsh-compact-basic` |
+| `command-compact` | `@deepseek-ai/dsh-command-compact` |
 | `subagent` | `@deepseek-ai/dsh-subagent` |
 | `subagent-spawn` | `@deepseek-ai/dsh-subagent-spawn` |
 | `subagent-fork` | `@deepseek-ai/dsh-subagent-fork` |
@@ -157,6 +182,17 @@ flowchart LR
 | `timeout-policy` | `@deepseek-ai/dsh-timeout-policy` |
 | `spill-local` | `@deepseek-ai/dsh-spill-local` |
 | `spill-policy` | `@deepseek-ai/dsh-spill-policy` |
+| `session-checkpoint-policy` | `@deepseek-ai/dsh-session-checkpoint-policy` |
+| `tool-result-prune` | `@deepseek-ai/dsh-compact-tool-result-prune` |
+| `tool-todo` | `@deepseek-ai/dsh-tool-todo` |
+| `tool-goal` | `@deepseek-ai/dsh-tool-goal` |
+| `tool-ralph` | `@deepseek-ai/dsh-tool-ralph` |
+| `tool-session-query` | `@deepseek-ai/dsh-tool-session-query` |
+| `tool-str-replace-editor` | `@deepseek-ai/dsh-tool-str-replace-editor` |
+| `repeat-tool-guard` | `@deepseek-ai/dsh-repeat-tool-guard` |
+| `web` | `@deepseek-ai/dsh-web` |
+| `web-search-deepseek` | `@deepseek-ai/dsh-web-search-deepseek` |
+| `tool-web` | `@deepseek-ai/dsh-tool-web` |
 | `tools` | `@deepseek-ai/dsh-tools` |
 | `system-prompt` | `@deepseek-ai/dsh-system-prompt` |
 | `agent-loop` | `@deepseek-ai/dsh-agent-loop` |

apps/cli/config/base.cordis.yml

@@ -204,6 +204,11 @@
 - id: compact-basic
   name: '@deepseek-ai/dsh-compact-basic'
 
+# Human `/compact`: one useful reduction below the automatic threshold. Backend
+# independent, so it follows whichever compaction service this leaf mounts.
+- id: command-compact
+  name: '@deepseek-ai/dsh-command-compact'
+
 - id: subagent
   name: '@deepseek-ai/dsh-subagent'
 
@@ -248,6 +253,74 @@
   config:
     maxInlineBytes: 50000
 
+# Durability checkpoints before each model request and top-level dispatch.
+- id: session-checkpoint-policy
+  name: '@deepseek-ai/dsh-session-checkpoint-policy'
+
+# Compacts oversized tool results before the broader conversation compactor
+# runs, preserving the model-visible result within the configured budget.
+- id: tool-result-prune
+  name: '@deepseek-ai/dsh-compact-tool-result-prune'
+  config:
+    thresholdChars: 8192
+    headChars: 4096
+    tailChars: 1024
+
+- id: tool-todo
+  name: '@deepseek-ai/dsh-tool-todo'
+
+# Persisted same-session goals reach the model and the slash menu here; the
+# domain, driver, and `/goal` command are above.
+- id: tool-goal
+  name: '@deepseek-ai/dsh-tool-goal'
+
+# Fresh-agent Ralph iteration over a build-time-fixed script.
+- id: tool-ralph
+  name: '@deepseek-ai/dsh-tool-ralph'
+  config:
+    subagentProvider: spawn
+    maxRounds: 64
+
+- id: tool-session-query
+  name: '@deepseek-ai/dsh-tool-session-query'
+  config:
+    maxSearchResults: 100
+    searchTimeoutMs: 30000
+
+- id: tool-str-replace-editor
+  name: '@deepseek-ai/dsh-tool-str-replace-editor'
+  config:
+    maxOutputChars: 16000
+
+# Consecutive-repeat reminders on the tool chain.
+- id: repeat-tool-guard
+  name: '@deepseek-ai/dsh-repeat-tool-guard'
+  config:
+    thresholds: [3, 5, 8]
+    argumentsPreviewChars: 500
+
+# Every surface enables the stable web_search model surface. DeepSeek search
+# resolves the same DEEPSEEK_API_KEY credential the Models page manages for
+# chat, at each search; its Messages endpoint is separate from the
+# chat-completions endpoint, so it takes its own base-URL override. Fetch stays
+# disabled and no fetch provider is mounted: that provider defers SSRF
+# protection and the model would choose the request target.
+- id: web
+  name: '@deepseek-ai/dsh-web'
+  config:
+    searchProvider: deepseek-official
+
+- id: web-search-deepseek
+  name: '@deepseek-ai/dsh-web-search-deepseek'
+  config:
+    apiKeyEnv: DEEPSEEK_API_KEY
+    baseURL: !!js process.env.DEEPSEEK_SEARCH_BASE_URL
+
+- id: tool-web
+  name: '@deepseek-ai/dsh-tool-web'
+  config:
+    fetch: false
+
 # ── rows every surface mounts, whose values each overlay states ──────────────
 
 # The tool registry. Presentation mode is a surface choice, so each overlay

apps/cli/config/core-web.cordis.yml

@@ -37,6 +37,20 @@
 - id: tool-todo
   disabled: true
 
+# These consumers are shared defaults on the ordinary shipped surfaces, but
+# this opt-in profile keeps exactly its two named tools.
+- id: tool-goal
+  disabled: true
+
+- id: tool-ralph
+  disabled: true
+
+- id: tool-session-query
+  disabled: true
+
+- id: tool-str-replace-editor
+  disabled: true
+
 # The matching browser controls must not offer host tools that this profile
 # omits. ui-question's host half owns the ask_user_question registration.
 - id: ui-plan

apps/cli/config/tui.cordis.yml

@@ -71,29 +71,17 @@
     - id: agent-loop-invariant
       name: '@deepseek-ai/dsh-agent-loop/invariant'
 
-    - id: session-checkpoint-policy
-      name: '@deepseek-ai/dsh-session-checkpoint-policy'
-
     # The derived query index behind `/resume`. The launcher provides a unique
     # process-local path because this SQLite backend has one writer owner; the
     # project-local fallback applies when no launcher sets the typed slot.
     - id: session-reference
       name: '@deepseek-ai/dsh-session-reference'
 
-    # Compacts oversized tool results before the broader conversation compactor
-    # runs, preserving the model-visible result within the configured budget.
-    - id: tool-result-prune
-      name: '@deepseek-ai/dsh-compact-tool-result-prune'
-
-    # Persisted same-session goals reach the model and the slash menu here; the
-    # domain, driver, and `/goal` command are in the base.
-    - id: tool-goal
-      name: '@deepseek-ai/dsh-tool-goal'
-
-    # A separate fixed consumer demonstrates fresh-agent Ralph iteration without
-    # changing the workflow tool or same-session goal behavior.
-    - id: tool-ralph
-      name: '@deepseek-ai/dsh-tool-ralph'
+    # Terminal-multiplexer context, mounted only where a terminal exists.
+    - id: tmux-context
+      name: '@deepseek-ai/dsh-tmux-context'
+      config:
+        refreshIntervalMs: 900000
 
     # The keyboard-backed provider behind ask_user_question and the plan-mode
     # review, and the front door it renders inside.

apps/cli/config/web.cordis.yml

@@ -80,28 +80,6 @@
 - id: fs-local
   disabled: true
 
-# The Web/headless product enables only the stable web_search model surface.
-# DeepSeek search resolves the same DEEPSEEK_API_KEY credential the Models page
-# manages for chat, at each search; its Messages endpoint is separate from the
-# chat-completions endpoint. Fetch remains disabled and no default fetch
-# provider is mounted.
-- insert:
-    - id: web
-      name: '@deepseek-ai/dsh-web'
-      config:
-        searchProvider: deepseek-official
-
-    - id: web-search-deepseek
-      name: '@deepseek-ai/dsh-web-search-deepseek'
-      config:
-        apiKeyEnv: DEEPSEEK_API_KEY
-        baseURL: !!js process.env.DEEPSEEK_SEARCH_BASE_URL
-
-    - id: tool-web
-      name: '@deepseek-ai/dsh-tool-web'
-      config:
-        fetch: false
-
 # ── web-only host rows, the transport layer, and the browser roster ─────────
 
 # `dshClient` rows are the browser roster the modules node half scans into
@@ -135,9 +113,6 @@
         writeEveryEvents: 200
         writeIntervalMs: 5000
 
-    - id: tool-todo
-      name: '@deepseek-ai/dsh-tool-todo'
-
     # Resolve bind host, SSH launch, and display once at boot, then mount the
     # matching dual-face directory picker. Mount -native or -browse directly in
     # an overlay to pin the interaction.

apps/cli/package.json

@@ -48,6 +48,7 @@
     "@deepseek-ai/dsh-client-ui-trajectory": "workspace:^",
     "@deepseek-ai/dsh-client-ui-workspace": "workspace:^",
     "@deepseek-ai/dsh-code-runtime-worker": "workspace:^",
+    "@deepseek-ai/dsh-command-compact": "workspace:^",
     "@deepseek-ai/dsh-command-goal": "workspace:^",
     "@deepseek-ai/dsh-commands": "workspace:^",
     "@deepseek-ai/dsh-compact-basic": "workspace:^",
@@ -73,6 +74,7 @@
     "@deepseek-ai/dsh-paths": "workspace:^",
     "@deepseek-ai/dsh-permission": "workspace:^",
     "@deepseek-ai/dsh-plan-mode": "workspace:^",
+    "@deepseek-ai/dsh-repeat-tool-guard": "workspace:^",
     "@deepseek-ai/dsh-pty": "workspace:^",
     "@deepseek-ai/dsh-pty-local": "workspace:^",
     "@deepseek-ai/dsh-sandbox-local": "workspace:^",
@@ -104,6 +106,7 @@
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
     "@deepseek-ai/dsh-tasks-local": "workspace:^",
     "@deepseek-ai/dsh-timeout-policy": "workspace:^",
+    "@deepseek-ai/dsh-tmux-context": "workspace:^",
     "@deepseek-ai/dsh-token-meter": "workspace:^",
     "@deepseek-ai/dsh-tool-ask-user": "workspace:^",
     "@deepseek-ai/dsh-tool-bash": "workspace:^",
@@ -113,6 +116,7 @@
     "@deepseek-ai/dsh-tool-fs-search": "workspace:^",
     "@deepseek-ai/dsh-tool-goal": "workspace:^",
     "@deepseek-ai/dsh-tool-ralph": "workspace:^",
+    "@deepseek-ai/dsh-tool-session-query": "workspace:^",
     "@deepseek-ai/dsh-tool-skill": "workspace:^",
     "@deepseek-ai/dsh-tool-str-replace-editor": "workspace:^",
     "@deepseek-ai/dsh-tool-subagent": "workspace:^",

apps/cli/tests/fixtures/composition-echo-llm.ts

@@ -0,0 +1,51 @@
+import type { Context } from 'cordis'
+import type {
+  GenerateOptions,
+  LlmModelInfo,
+  LlmResolvedModelInfo,
+  StreamChunk,
+} from '@deepseek-ai/dsh-llm'
+import { LlmAdapter } from '@deepseek-ai/dsh-llm'
+
+/** Terminal marker the preset smoke waits for before it asks the TUI to exit. */
+export const COMPOSITION_REPLY_TEXT = 'Shipped composition acknowledged.'
+
+// Provider id and model the keyless tail routes `main` to; that overlay is the
+// only caller, so the pair lives here as plain constants.
+const COMPOSITION_PROVIDER = 'composition-keyless'
+const COMPOSITION_MODEL = 'composition-keyless-model'
+
+/**
+ * Network-free adapter for the shipped-composition smoke. It answers every
+ * request — tool-ful agent turns and the tool-less auxiliary calls alike — with
+ * one fixed text and never calls a tool, because the assertion under test is the
+ * assembled tool catalog the loop logs, not any tool's behavior.
+ */
+class CompositionEchoAdapter extends LlmAdapter {
+  override listModels(provider: string): Promise<readonly LlmModelInfo[]> {
+    return Promise.resolve([{ provider, id: COMPOSITION_MODEL, name: 'Preset Keyless' }])
+  }
+
+  override resolveModel(provider: string, model: string): Promise<LlmResolvedModelInfo> {
+    return Promise.resolve({ provider, id: model, name: 'Preset Keyless', context: { contextWindow: 128_000 } })
+  }
+
+  override async * stream(_options: GenerateOptions): AsyncIterable<StreamChunk> {
+    yield { type: 'block-start', index: 0, blockType: 'text' }
+    for (const char of COMPOSITION_REPLY_TEXT) yield { type: 'text-delta', index: 0, text: char }
+    yield { type: 'block-end', index: 0, block: { type: 'text', text: COMPOSITION_REPLY_TEXT } }
+    yield { type: 'usage', usage: { inputTokens: 20, outputTokens: COMPOSITION_REPLY_TEXT.length } }
+    yield { type: 'finish', reason: { kind: 'stop' } }
+  }
+}
+
+export const name = 'composition-echo-llm'
+export const inject = ['llm']
+
+/**
+ * Register the network-free adapter the shipped-composition smoke routes through.
+ * @param ctx - the loader-mounted plugin context.
+ */
+export function apply(ctx: Context): void {
+  ctx.llm.registerAdapter([COMPOSITION_PROVIDER], new CompositionEchoAdapter())
+}

apps/cli/tests/fixtures/composition-keyless-tail.cordis.yml

@@ -0,0 +1,52 @@
+# Keyless tail for the shipped-composition smoke, applied as `--config` so the
+# launcher boots `base.cordis.yml` + `tui.cordis.yml` and then this file.
+#
+# Everything below is test isolation, never composition under test: the model is
+# replaced so no request leaves the process, the settle marker gates the smoke's
+# first prompt, and the session artifacts move into the smoke's temporary
+# workspace so the log inspection can read them.
+
+# A patch's `name` is an assertion rather than a replacement, so the base
+# adapter row is disabled and the scripted one inserted. Relative specifiers
+# resolve against the INCLUDED file's directory (apps/cli/config), not this
+# file's, because the include moves baseUrl there.
+- id: llm-deepseek
+  disabled: true
+
+- insert:
+    - id: composition-echo-llm
+      name: '../tests/fixtures/composition-echo-llm.ts'
+    - id: composition-settled
+      name: '../tests/fixtures/composition-settled.ts'
+
+- id: agent-loop
+  config:
+    agents:
+      - id: main
+        provider: composition-keyless
+        model: composition-keyless-model
+        cwd: !!js process.cwd()
+
+- id: session-persistence-jsonl
+  config:
+    root: './.sessions'
+    compression: none
+
+- id: session-query-sqlite
+  config:
+    path: './.sessions/session-query.db'
+
+# The title call is a second, tool-less request that would race the log
+# inspection for no coverage: the catalog under test rides the agent turn.
+- id: session-title-llm
+  disabled: true
+
+- id: tui
+  config:
+    sessionId: !!js configuredAgentIdentities?.main?.id ?? 'main'
+    welcome: 'composition smoke ready.'
+    showReasoning: true
+
+# HMR watches the repository; a PTY subprocess test must not start a watcher.
+- id: hmr
+  disabled: true

apps/cli/tests/fixtures/composition-settled.ts

@@ -0,0 +1,24 @@
+import type { Context } from 'cordis'
+
+/**
+ * Marker the shipped-composition smoke gates its first prompt on. The TUI renders as soon as
+ * its own fiber starts, so a prompt typed at the banner can reach the loop while
+ * later rows — tool plugins, persistence — are still activating, and would
+ * assemble a partial catalog. Waiting for this line makes the turn observe the
+ * settled tree.
+ */
+export const COMPOSITION_SETTLED_MARKER = 'COMPOSITION_TREE_SETTLED'
+
+export const name = 'composition-settled'
+
+/**
+ * Announce settled Loader activation on the terminal byte stream, after every
+ * entry in the booted tree has started. The write is detached: awaiting the
+ * Loader from inside an entry would wait on this entry's own activation.
+ * @param ctx - the loader-mounted plugin context.
+ */
+export function apply(ctx: Context): void {
+  void ctx.loader.await().then(() => {
+    process.stdout.write(`\n${COMPOSITION_SETTLED_MARKER}\n`)
+  })
+}

apps/cli/tests/shipped-composition.e2e.ts

@@ -0,0 +1,124 @@
+import { readdir, readFile } from 'node:fs/promises'
+import { fileURLToPath } from 'node:url'
+import { join } from 'node:path'
+import { describe, expect, it } from 'vitest'
+import { LOADER_SMOKE_TEST_TIMEOUT_MS } from '@deepseek-ai/dsh-loader-smoke'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import { COMPOSITION_REPLY_TEXT } from './fixtures/composition-echo-llm.ts'
+import { COMPOSITION_SETTLED_MARKER } from './fixtures/composition-settled.ts'
+import { runTuiPtySmoke } from './pty-harness.ts'
+
+const dshBinScript = fileURLToPath(new URL('../src/bin.ts', import.meta.url))
+const tsconfigPath = fileURLToPath(new URL('../../../tsconfig.json', import.meta.url))
+// An overlay over the shipped tree, so the catalog under test is the one
+// `base.cordis.yml` + `tui.cordis.yml` assemble; the tail only swaps the model
+// and redirects session artifacts.
+const keylessTail = fileURLToPath(new URL('./fixtures/composition-keyless-tail.cordis.yml', import.meta.url))
+
+/**
+ * The catalog the shipped `dsh` TUI puts in front of the model, as the loop
+ * logged it, minus the ripgrep-dependent pair below.
+ * The absences are the composition's security decisions, not incidental gaps:
+ * the `cordis_*` toolset executes model-written JavaScript that no sandbox row
+ * confines, `web_fetch` chooses its own request target, and `mcp_*` servers
+ * spawn outside `ctx.bash`. The composition Agent Note owns the rationale and
+ * its sources.
+ */
+const EXPECTED_TUI_TOOLS = [
+  'ask_user_question',
+  'bash',
+  'create_goal',
+  'edit',
+  'exit_plan_mode',
+  'get_goal',
+  'ralph',
+  'read',
+  'session_event_read',
+  'session_event_search',
+  'session_event_trace',
+  'session_search',
+  'session_trace',
+  'skill',
+  'str_replace_editor',
+  'subagent',
+  'subagent_fork',
+  'task_kill',
+  'task_list',
+  'task_output',
+  'todo_write',
+  'update_goal',
+  'web_search',
+  'workflow',
+  'write',
+]
+
+/**
+ * `glob` and `grep` come from `dsh-tool-fs-search`, which probes `command -v rg`
+ * through the mounted bash executor at load and registers neither tool when
+ * ripgrep is absent. That is a host dependency, not a composition decision, so the
+ * pair is asserted separately — present together or absent together.
+ */
+const RIPGREP_TOOLS = ['glob', 'grep']
+
+/** The assembled request header the smoke asserts on. */
+interface LoggedHeader {
+  /** Assembled tool names, sorted. */
+  names: string[]
+  /** `bash`'s assembled parameter properties; the escalation pair is present only under a confining executor. */
+  bashArguments: Record<string, unknown>
+}
+
+/**
+ * Read the request header the loop assembled for its first request from the
+ * session log the smoke's workspace persisted — the model-visible composition
+ * itself, not a registry projection taken beside it.
+ * @param cwd - the smoke's temporary workspace.
+ * @returns the assembled catalog, system prompt, and `bash` argument shape.
+ */
+async function loggedHeader(cwd: string): Promise<LoggedHeader> {
+  const sessionsDir = join(cwd, '.sessions')
+  const entries = await readdir(sessionsDir, { recursive: true })
+  // A single keyless run writes one session log.
+  const logRelPath = entries.find(name => name.endsWith('.jsonl'))
+  if (logRelPath === undefined) throw new Error(`no session log written under ${sessionsDir}`)
+  const lines = (await readFile(join(sessionsDir, logRelPath), 'utf8')).split('\n').filter(Boolean)
+  for (const line of lines) {
+    const event = JSON.parse(line) as SessionEvent
+    if (event.type !== 'request/header') continue
+    const tools = event.data.header.tools ?? []
+    const bash = tools.find(schema => schema.name === 'bash')
+    return {
+      names: tools.map(schema => schema.name).sort(),
+      bashArguments: (bash?.parameters as { properties?: Record<string, unknown> } | undefined)?.properties ?? {},
+    }
+  }
+  throw new Error(`session log ${logRelPath} has no request/header event`)
+}
+
+describe('shipped dsh composition (real Loader tree in a PTY)', () => {
+  it('assembles exactly the shipped TUI catalog', async () => {
+    let observed: LoggedHeader | undefined
+    const output = await runTuiPtySmoke({
+      label: 'dsh shipped composition',
+      tempDirPrefix: 'dsh-shipped-tui-',
+      binScript: dshBinScript,
+      tsconfigPath,
+      configPath: keylessTail,
+      env: { DEEPSEEK_API_KEY: 'keyless-composition-no-call', DSH_TELEMETRY_DISABLED: '1' },
+      // Artifact CI builds and smokes concurrently on a contended runner.
+      ...(process.env.DSH_EXAMPLE_MODE === 'lib' ? { timeoutMs: 60_000 } : {}),
+      actions: [
+        { waitFor: COMPOSITION_SETTLED_MARKER, send: 'Describe the shipped composition.\r' },
+        { waitFor: COMPOSITION_REPLY_TEXT, send: '/exit\r' },
+      ],
+      inspect: async (cwd) => { observed = await loggedHeader(cwd) },
+    })
+    expect(output).toContain(COMPOSITION_REPLY_TEXT)
+    expect(observed?.names.filter(name => !RIPGREP_TOOLS.includes(name))).toEqual(EXPECTED_TUI_TOOLS)
+    expect([[], RIPGREP_TOOLS]).toContainEqual(observed?.names.filter(name => RIPGREP_TOOLS.includes(name)))
+    // The TUI mounts the unrestricted local executors, so `tool-bash` emits no
+    // escalation pair. Pinning its absence keeps a later sandbox change from
+    // arriving here unannounced.
+    expect(Object.keys(observed?.bashArguments ?? {})).not.toContain('sandbox_permissions')
+  }, LOADER_SMOKE_TEST_TIMEOUT_MS)
+})

apps/cli/tests/snapshots/queued-manual-compact/terminal.expected.txt

@@ -0,0 +1,128 @@
+terminal 100x36 buffer=normal length=66 base=30 viewport=30
+lifecycle started=1 stopped=0 progress=inactive
+title "Reply with exactly the word: — DSH TUI snapshot"
+cursor hidden column=7 viewportRow=35 bufferRow=65
+buffer
+0| " DEEPSEEK HARNESS"
+  style 1-8 fg=bright-magenta bold
+  style 10-16 bold
+1| " Reply with exactly the word:"
+  style 1-28 dim
+2| " main-session"
+  style 1-12 dim
+3| <blank>
+4| "Context · snapshot-seed"
+  style 0-22 dim
+5| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+6| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+7| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+8| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+9| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+10| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+11| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+12| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+13| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+14| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+15| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+16| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+17| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+18| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-99 dim
+19| "Older snapshot context. Older snapshot context. Older snapshot context. Older snapshot context.     "
+  style 0-94 dim
+20| <blank>
+21| "You                                                                                                 "
+  style 0-2 fg=bright-magenta bold underline
+22| "Reply with exactly the word: ONE. No tools.                                                         "
+23| <blank>
+24| "Assistant                                                                                           "
+  style 0-8 fg=bright-magenta bold underline
+25| "Reasoning                                                                                           "
+  style 0-8 dim italic
+26| "The user wants me to reply with exactly the word \"ONE\" and use no tools.                            "
+  style 0-71 dim italic
+27| "ONE                                                                                                 "
+28| "Model wait 0.0s · Completed 2026-07-21 12:00:00                                                     "
+  style 0-46 dim
+29| <blank>
+30| "Keyboard shortcuts                                                                                  "
+  style 0-17 fg=bright-magenta bold
+31| "Enter send • Shift/Alt+Enter newline • Up/Down prompt history                                       "
+  style 0-60 dim
+32| "Esc cancel turn • Ctrl+O cycle cards (collapse/expand/hide) • Ctrl+R toggle reasoning • Ctrl+L      "
+  style 0-99 dim
+33| "redraw                                                                                              "
+  style 0-5 dim
+34| "Ctrl+C cancel while running; clear input or exit while idle • Ctrl+D exit                           "
+  style 0-72 dim
+35| "                                                                                                    "
+36| "/clear — Clear the transcript view (session history is unchanged)                                   "
+  style 0-64 dim
+37| "/compact — Compact older conversation history                                                       "
+  style 0-44 dim
+38| "/exit — Exit after the active turn reaches idle                                                     "
+  style 0-46 dim
+39| "/help — Show keyboard shortcuts and commands                                                        "
+  style 0-43 dim
+40| "/model [[provider/]model] — Show or switch this session's model                                     "
+  style 0-62 dim
+41| "/palette — Show every color and attribute role this terminal renders                                "
+  style 0-67 dim
+42| "/quit — Exit after the active turn reaches idle                                                     "
+  style 0-46 dim
+43| "/reload — EXPERIMENTAL (dev): re-read loader config files and apply the diff (idle only)            "
+  style 0-87 dim
+44| "/resume — List this workspace's resumable sessions                                                  "
+  style 0-49 dim
+45| "/status — Show session diagnostics, system prompt, and registered tools                             "
+  style 0-70 dim
+46| "/skill:<name> [instructions] — load a skill into the conversation                                   "
+  style 0-64 dim
+47| <blank>
+48| "Context · snapshot-injector"
+  style 0-26 dim
+49| "Injected while compaction was running.                                                              "
+  style 0-37 dim
+50| <blank>
+51| "… earlier context was compacted …                                                                   "
+  style 0-32 dim
+52| <blank>
+53| "You                                                                                                 "
+  style 0-2 fg=bright-magenta bold underline
+54| "Reply with exactly the word: TWO. No tools.                                                         "
+55| <blank>
+56| "Compacted 2 history items (~387 tokens).                                                            "
+  style 0-39 dim
+57| <blank>
+58| "Assistant                                                                                           "
+  style 0-8 fg=bright-magenta bold underline
+59| "Reasoning                                                                                           "
+  style 0-8 dim italic
+60| "The user wants me to reply with exactly the word \"TWO\" and no tools.                                "
+  style 0-67 dim italic
+61| "TWO                                                                                                 "
+62| "Model wait 0.0s · Completed 2026-07-21 12:00:00                                                     "
+  style 0-46 dim
+63| <blank>
+64| "/workspace/project  deepseek-v4-flash  ↑2.9k ↓41  cache 49%  3% cont"
+  style 0-49 fg=bright-magenta bold
+  style 52-68 dim
+  style 71-90 dim
+  style 93-99 dim
+65| " dsh ◍                                                                                              "
+  style 1-3 fg=bright-magenta bold
+  style 5-6 dim
+  style 7-7 inverse

apps/cli/tests/tui.snapshot.ts

@@ -11,8 +11,12 @@ import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import LocalSubprocessService from '@deepseek-ai/dsh-subprocess-local'
 import WorkerCodeRuntime from '@deepseek-ai/dsh-code-runtime-worker'
 import CommandService from '@deepseek-ai/dsh-commands'
+import * as CommandCompact from '@deepseek-ai/dsh-command-compact'
+import { BasicCompactService } from '@deepseek-ai/dsh-compact-basic'
+import type { SummarizationInput } from '@deepseek-ai/dsh-compact-basic/src/summarizer.ts'
 import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
 import * as FsPolicy from '@deepseek-ai/dsh-fs-policy'
+import { createUserMessage } from '@deepseek-ai/dsh-llm'
 import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
 import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
 import { installLlmReplay, parseSessionLog } from '@deepseek-ai/dsh-llm-replay'
@@ -45,6 +49,8 @@ type ScenarioInteraction = 'skill-invocation-policy'
 
 interface Scenario {
   name: string
+  /** Replay fixture owned by an earlier scenario, for a derived presentation case. */
+  fixture?: string
   composition: Composition
   expectedTools: string[]
   expectedEventCounts?: Record<string, number>
@@ -68,6 +74,13 @@ interface Scenario {
   spillMaxInlineBytes?: number
   /** Run scenario-specific terminal input instead of replaying recorded user prompts. */
   interaction?: ScenarioInteraction
+  /**
+   * Mount a deterministic compaction backend plus `/compact`, then run the
+   * human command with a held summary while a prompt and injected context
+   * arrive. Proves queued input waits for the standalone bracket's durability
+   * checkpoint instead of racing the replacement.
+   */
+  manualCompact?: boolean
 }
 
 const SCENARIOS: Scenario[] = [
@@ -80,6 +93,14 @@ const SCENARIOS: Scenario[] = [
     leavePlanModeAfterFirstTurn: true,
     recorded: true,
   },
+  {
+    name: 'queued-manual-compact',
+    fixture: 'multi-turn-conversation',
+    composition: 'native',
+    expectedTools: [],
+    recorded: false,
+    manualCompact: true,
+  },
   {
     name: 'todo-plan',
     composition: 'native',
@@ -149,6 +170,44 @@ function snapshotModeFromEnv(value: string | undefined): SnapshotMode {
 
 const MODE = snapshotModeFromEnv(process.env.DSH_SNAPSHOT)
 const observedScenarios = new Set<string>()
+const workerState = Reflect.get(globalThis, '__vitest_worker__') as
+  | { readonly config?: { readonly testNamePattern?: RegExp } }
+  | undefined
+// Worker argv omits the parent CLI's `-t`; the serialized runner config is the
+// authoritative distinction between a focused replay and the full suite.
+const TEST_NAME_FILTERED = workerState?.config?.testNamePattern !== undefined
+
+/**
+ * Deterministic keyless summary that pauses so the scenario can submit a real
+ * prompt and inject context while manual compaction holds turn admission.
+ */
+class DeferredSnapshotCompactService extends BasicCompactService {
+  readonly summaryStarted = Promise.withResolvers<undefined>()
+  readonly releaseSummary = Promise.withResolvers<undefined>()
+
+  override async summarize(
+    _input: SummarizationInput,
+    _agent: Agent,
+    signal?: AbortSignal,
+  ): Promise<{ summary: [{ type: 'text'; text: string }]; provider: string; model: string }> {
+    this.summaryStarted.resolve(undefined)
+    await this.releaseSummary.promise
+    signal?.throwIfAborted()
+    return {
+      summary: [{ type: 'text', text: 'Keyless manual compaction checkpoint.' }],
+      provider: 'snapshot',
+      model: 'snapshot-compactor',
+    }
+  }
+}
+
+/** Seed between-turn model-visible history without inventing a loop execution. */
+function seedCompactableHistory(agent: Agent): void {
+  agent.inject(createUserMessage({
+    content: [{ type: 'text', text: 'Older snapshot context. '.repeat(60) }],
+    source: { kind: 'plugin', plugin: 'snapshot-seed' },
+  }))
+}
 
 function snapshotDisplayPath(displayPath: string, cwd: string, displayCwd: string): string {
   const rel = relative(cwd, displayPath)
@@ -161,10 +220,15 @@ function scenarioDir(scenario: Scenario): string {
   return join(SNAPSHOTS_DIR, scenario.name)
 }
 
+/** Directory owning the replay fixture: the scenario's own, or the one it derives from. */
+function fixtureDir(scenario: Scenario): string {
+  return join(SNAPSHOTS_DIR, scenario.fixture ?? scenario.name)
+}
+
 function childFixturePaths(scenario: Scenario): string[] {
   return Array.from(
     { length: scenario.childSessions ?? 0 },
-    (_, index) => join(scenarioDir(scenario), `session.${index + 1}.jsonl`),
+    (_, index) => join(fixtureDir(scenario), `session.${index + 1}.jsonl`),
   )
 }
 
@@ -206,6 +270,24 @@ async function settleTerminal(terminal: HeadlessTerminal): Promise<void> {
   if (stable < 3) throw new Error('TUI frames did not quiesce within 200ms')
 }
 
+/** Bound deterministic in-process coordination waits with actionable state. */
+async function snapshotDeadline<T>(
+  operation: Promise<T>,
+  detail: () => string,
+): Promise<T> {
+  let timer: ReturnType<typeof setTimeout> | undefined
+  try {
+    return await Promise.race([
+      operation,
+      new Promise<never>((_resolve, reject) => {
+        timer = setTimeout(() => { reject(new Error(detail())) }, 5_000)
+      }),
+    ])
+  } finally {
+    if (timer !== undefined) clearTimeout(timer)
+  }
+}
+
 async function mountScenarioContext(
   scenario: Scenario,
   cwd: string,
@@ -232,6 +314,9 @@ async function mountScenarioContext(
     skills: { local: { agentsHome: join(cwd, '.agents') } },
   })
   await ctx.plugin(TokenMeterService)
+  if (scenario.manualCompact === true) {
+    await ctx.plugin(DeferredSnapshotCompactService, { auto: false })
+  }
   await ctx.plugin(LocalSubprocessService)
   await ctx.plugin(LocalBashExecutor, { cwd, timeoutMs: 30_000 })
   await ctx.plugin(SnapshotLocalFileSystem, { cwd: '/' })
@@ -249,6 +334,7 @@ async function mountScenarioContext(
   await ctx.plugin(ToolWorkflow)
   await ctx.plugin(ToolRalph)
   await ctx.plugin(CommandService)
+  if (scenario.manualCompact === true) await ctx.plugin(CommandCompact)
   if (scenario.enterPlanMode === true) {
     await ctx.plugin(PlanModeService, { section: 'Snapshot plan mode instructions.' })
   }
@@ -276,9 +362,9 @@ interface ScenarioResult {
 }
 
 async function runScenario(scenario: Scenario): Promise<ScenarioResult> {
-  const clock = vi.spyOn(Date, 'now').mockReturnValue(new Date(2026, 6, 21, 12, 0, 0).getTime())
-  const dir = scenarioDir(scenario)
-  const fixtureFile = join(dir, 'session.jsonl')
+  const snapshotTime = new Date(2026, 6, 21, 12, 0, 0).getTime()
+  const clock = vi.spyOn(Date, 'now').mockReturnValue(snapshotTime)
+  const fixtureFile = join(fixtureDir(scenario), 'session.jsonl')
   const childFiles = childFixturePaths(scenario)
   const prompts = userPrompts(await readFile(fixtureFile, 'utf8'))
   if (scenario.interaction === undefined) {
@@ -292,7 +378,7 @@ async function runScenario(scenario: Scenario): Promise<ScenarioResult> {
   const terminal = new HeadlessTerminal(100, 36)
   try {
     if (scenario.seedWorkspace === true) {
-      const source = join(scenarioDir(scenario), 'workspace')
+      const source = join(fixtureDir(scenario), 'workspace')
       await cp(source, cwd, { recursive: true })
     }
     ctx = await mountScenarioContext(scenario, cwd, displayCwd, fixtureFile, childFiles)
@@ -308,6 +394,7 @@ async function runScenario(scenario: Scenario): Promise<ScenarioResult> {
       agentOptions: { provider: 'deepseek-official', model: 'deepseek-v4-flash' },
     })
     const agent: Agent = handle.agent
+    if (scenario.manualCompact === true) seedCompactableHistory(agent)
     controller = createTuiChat(ctx, {
       sessionId: 'main-session',
       theme: { color: true },
@@ -380,6 +467,14 @@ async function runScenario(scenario: Scenario): Promise<ScenarioResult> {
     }
 
     let remainingPrompts = prompts
+    let queuedPrompt: string | undefined
+    let manualOrder: string[] | undefined
+    let manualCommandId: string | undefined
+    if (scenario.manualCompact === true) {
+      expect(prompts.length, 'queued manual compaction needs a second replayed prompt').toBeGreaterThanOrEqual(2)
+      queuedPrompt = prompts.at(-1)
+      remainingPrompts = prompts.slice(0, -1)
+    }
     if (scenario.enterPlanMode === true) {
       const firstPrompt = prompts[0]!
       terminal.send(`/plan ${firstPrompt}`)
@@ -396,12 +491,93 @@ async function runScenario(scenario: Scenario): Promise<ScenarioResult> {
     }
 
     for (const prompt of remainingPrompts) {
+      const admitted = agent.session.events.filter(event =>
+        event.type === 'user/message' && event.data.source.kind === 'user').length
       terminal.send(prompt)
       terminal.send('\r')
+      await terminal.flush()
+      await expect.poll(() => agent.session.events.filter(event =>
+        event.type === 'user/message' && event.data.source.kind === 'user').length).toBe(admitted + 1)
       await agent.whenIdle()
       await settleTerminal(terminal)
     }
 
+    if (scenario.manualCompact === true && queuedPrompt !== undefined) {
+      terminal.send('/help')
+      terminal.send('\r')
+      await settleTerminal(terminal)
+      expect(await terminal.snapshot({ includeScrollback: true }))
+        .toContain('/compact — Compact older conversation history')
+
+      const compact = ctx.compact as DeferredSnapshotCompactService
+      const inbox: string[] = []
+      manualOrder = []
+      ctx.on('agent/inbox/enqueue', (subject, item) => {
+        if (subject === agent) inbox.push(`enqueue:${item.placement}:${item.id}`)
+      })
+      ctx.on('agent/inbox/dequeue', (subject, message) => {
+        if (subject === agent) inbox.push(`dequeue:${message.id}`)
+      })
+      ctx.on('session/event', (session, event) => {
+        if (session !== agent.session) return
+        if (event.type === 'command/run' && event.data.name === 'compact') {
+          manualCommandId = event.data.commandId
+          manualOrder?.push('command/run')
+        }
+        if (event.type === 'command/done' && event.data.commandId === manualCommandId) {
+          manualOrder?.push('command/done')
+        }
+        if (event.type.startsWith('compact/')) manualOrder?.push(event.type)
+        if (event.type === 'user/message'
+          && event.data.source.kind === 'plugin'
+          && event.data.source.plugin === 'compact') manualOrder?.push('checkpoint')
+        if (event.type === 'turn/start') manualOrder?.push(`turn/start:${event.data.trigger.kind}`)
+      })
+      ctx.on('session/flush', (session) => {
+        if (session === agent.session) manualOrder?.push('flush')
+      })
+
+      terminal.send('/compact')
+      terminal.send('\r')
+      await terminal.flush()
+      await snapshotDeadline(compact.summaryStarted.promise, () =>
+        `manual summary did not start; status=${agent.status}; tail=${
+          agent.session.events.slice(-8).map(event => event.type).join(',')
+        }`)
+      clock.mockReturnValue(snapshotTime + 1_000)
+      await settleTerminal(terminal)
+      await expect.poll(() => terminal.snapshot()).toContain('dsh ⊙')
+      await expect.poll(() => terminal.snapshot()).toContain('Context being compacted 1.0s')
+      const liveCompaction = await terminal.snapshot()
+      expect(liveCompaction.indexOf('Context being compacted 1.0s')).toBeLessThan(liveCompaction.indexOf('dsh ⊙'))
+      clock.mockReturnValue(snapshotTime)
+
+      // Real keystrokes: the prompt keeps its ordinary queue identity while
+      // admission is reserved, and an injection appends immediately.
+      terminal.send(queuedPrompt)
+      terminal.send('\r')
+      await terminal.flush()
+      await expect.poll(() => inbox.length).toBe(1)
+      agent.inject(createUserMessage({
+        content: [{ type: 'text', text: 'Injected while compaction was running.' }],
+        source: { kind: 'plugin', plugin: 'snapshot-injector' },
+      }))
+      expect(inbox[0]).toMatch(/^enqueue:queued:/u)
+      expect(agent.status).toBe('idle')
+      expect(agent.session.events.some(event => event.type === 'user/message'
+        && event.data.source.kind === 'user'
+        && event.data.content.some(block => block.type === 'text' && block.text === queuedPrompt))).toBe(false)
+
+      const idle = agent.whenIdle()
+      compact.releaseSummary.resolve(undefined)
+      await snapshotDeadline(idle, () =>
+        `manual compaction did not reach idle; status=${agent.status}; order=${manualOrder?.join(',') ?? ''}; tail=${
+          agent.session.events.slice(-12).map(event => event.type).join(',')
+        }`)
+      await settleTerminal(terminal)
+      expect(inbox).toEqual([inbox[0], `dequeue:${inbox[0]?.slice('enqueue:queued:'.length) ?? ''}`])
+    }
+
     const events: SessionEvent[] = [...agent.session.events]
     const firstHeader = events.find(event => event.type === 'request/header')
     expect(firstHeader?.type === 'request/header' && firstHeader.data.header.system)
@@ -437,6 +613,87 @@ async function runScenario(scenario: Scenario): Promise<ScenarioResult> {
       expect(events.filter(event => event.type === 'user/message' && event.data.source.kind === 'plugin').map(event => (event.data as { content: unknown }).content))
         .toContainEqual([{ type: 'text', text: 'The user switched this session back to the default mode.' }])
     }
+    if (scenario.manualCompact === true) {
+      const compactStart = events.find(event => event.type === 'compact/start')
+      const compactSummary = events.find(event => event.type === 'compact/summary')
+      const compactCheckpoint = events.find(event => event.type === 'user/message'
+        && event.data.source.kind === 'plugin' && event.data.source.plugin === 'compact')
+      const injectedEvent = events.find(event => event.type === 'user/message'
+        && event.data.source.kind === 'plugin' && event.data.source.plugin === 'snapshot-injector')
+      const compactEnd = events.find(event => event.type === 'compact/end')
+      expect(compactStart?.data.turn).toBeNull()
+      expect(compactEnd?.data.turn).toBeNull()
+      expect(events.filter(event => event.type === 'compact/summary')).toHaveLength(1)
+      if (compactStart === undefined || compactSummary === undefined
+        || compactCheckpoint === undefined || injectedEvent === undefined
+        || compactEnd === undefined) {
+        throw new Error('manual compaction snapshot is missing its durable marker, summary, checkpoint, or injection')
+      }
+      // The markers are time points, not an exclusive container: unrelated
+      // idle injection is allowed between them while the selected span stays stable.
+      expect(compactStart.seq).toBeLessThan(injectedEvent.seq)
+      expect(injectedEvent.seq).toBeLessThan(compactSummary.seq)
+      expect(compactSummary.seq).toBeLessThan(compactCheckpoint.seq)
+      expect(compactCheckpoint.seq).toBeLessThan(compactEnd.seq)
+
+      const manualTimeline = manualOrder ?? []
+      const commandRunIndex = manualTimeline.indexOf('command/run')
+      const compactStartIndex = manualTimeline.indexOf('compact/start')
+      const compactEndIndex = manualTimeline.indexOf('compact/end')
+      const firstFlushIndex = manualTimeline.indexOf('flush')
+      const queuedTurnIndex = manualTimeline.indexOf('turn/start:message')
+      const commandDoneIndex = manualTimeline.indexOf('command/done')
+      expect(manualTimeline.filter(item => item === 'command/run')).toHaveLength(1)
+      expect(manualTimeline.filter(item => item === 'command/done')).toHaveLength(1)
+      expect(compactStartIndex).toBeGreaterThan(commandRunIndex)
+      expect(compactEndIndex).toBeGreaterThan(compactStartIndex)
+      expect(firstFlushIndex).toBeGreaterThan(compactEndIndex)
+      expect(queuedTurnIndex).toBeGreaterThan(firstFlushIndex)
+      expect(commandDoneIndex).toBeGreaterThan(firstFlushIndex)
+
+      const commandRun = events.find(event => event.type === 'command/run'
+        && event.data.name === 'compact')
+      const commandRunId = commandRun?.type === 'command/run'
+        ? commandRun.data.commandId
+        : undefined
+      const commandDone = events.find(event => event.type === 'command/done'
+        && event.data.commandId === commandRunId)
+      expect(commandRun?.type === 'command/run' && commandRun.data).toEqual({
+        commandId: commandRunId,
+        name: 'compact',
+        args: '',
+        source: { kind: 'user' },
+      })
+      expect(commandDone?.type === 'command/done' && commandDone.data).toEqual({
+        commandId: commandRunId,
+        kind: 'success',
+        text: 'Compacted 2 history items (~387 tokens).',
+      })
+      expect(commandRun !== undefined && commandRun.seq < compactStart.seq).toBe(true)
+      expect(commandDone !== undefined && commandDone.seq > compactEnd.seq).toBe(true)
+      expect(agent.session.surface.nodes).not.toContain(commandRun?.seq)
+      expect(agent.session.surface.nodes).not.toContain(commandDone?.seq)
+
+      // The command line itself never becomes a prompt.
+      expect(events.some(event => event.type === 'user/message'
+        && event.data.source.kind === 'user'
+        && event.data.content.some(block => block.type === 'text' && block.text.trim() === '/compact'))).toBe(false)
+      const derived = agent.session.deriveMessages().map(message => message.content
+        .map(block => block.type === 'text' ? block.text : '')
+        .join(''))
+      const checkpoint = derived.findIndex(text => text.includes('Keyless manual compaction checkpoint.'))
+      const injected = derived.findIndex(text => text.includes('Injected while compaction was running.'))
+      const queued = derived.findIndex(text => text === queuedPrompt)
+      expect(checkpoint).toBe(0)
+      expect(injected).toBeGreaterThan(checkpoint)
+      expect(queued).toBeGreaterThan(injected)
+      expect(derived).not.toContain('/compact')
+      expect(derived).not.toContain('Compacted 2 history items (~387 tokens).')
+      expect(derived.filter(text => text.includes('Injected while compaction was running.'))).toHaveLength(1)
+      expect(compactSummary.data.shadowedSeqs).not.toContain(injectedEvent.seq)
+      const queuedTurn = events.findLast(event => event.type === 'turn/start')
+      expect(queuedTurn !== undefined && compactEnd.seq < queuedTurn.seq).toBe(true)
+    }
     if (scenario.spillMaxInlineBytes !== undefined) {
       // The REAL pipeline ran (tools execute on replay too): the durable
       // dispatch copy is bounded to a preview + locator under the run cwd,
@@ -514,7 +771,23 @@ describe('TUI recorded-session terminal snapshots', () => {
 })
 
 afterAll(async () => {
-  expect([...observedScenarios].sort()).toEqual(SCENARIOS.map(scenario => scenario.name).sort())
+  const scenarioNames = SCENARIOS.map(scenario => scenario.name).sort()
+  const observedNames = [...observedScenarios].sort()
+  if (TEST_NAME_FILTERED) {
+    expect(observedNames).not.toHaveLength(0)
+    expect(scenarioNames).toEqual(expect.arrayContaining(observedNames))
+  } else {
+    expect(observedNames).toEqual(scenarioNames)
+  }
+  for (const [index, scenario] of SCENARIOS.entries()) {
+    if (scenario.fixture === undefined) continue
+    const sourceIndex = SCENARIOS.findIndex(candidate => candidate.name === scenario.fixture)
+    expect(sourceIndex, `${scenario.name} fixture source ${scenario.fixture} must exist`).toBeGreaterThanOrEqual(0)
+    expect(sourceIndex, `${scenario.name} fixture source must precede it`).toBeLessThan(index)
+    const source = SCENARIOS[sourceIndex]
+    expect(source?.fixture, `${scenario.name} fixture source must own its replay files`).toBeUndefined()
+    expect(source?.recorded, `${scenario.name} fixture source must be recordable`).toBe(true)
+  }
   const directories = (await readdir(SNAPSHOTS_DIR, { withFileTypes: true }))
     .filter(entry => entry.isDirectory())
     .map(entry => entry.name)
@@ -522,14 +795,14 @@ afterAll(async () => {
   expect(directories).toEqual(SCENARIOS.map(scenario => scenario.name).sort())
   for (const scenario of SCENARIOS) {
     const expected = [
-      'session.jsonl',
+      ...scenario.fixture === undefined ? ['session.jsonl'] : [],
       'terminal.expected.txt',
-      ...scenario.seedWorkspace === true ? ['workspace'] : [],
+      ...scenario.seedWorkspace === true && scenario.fixture === undefined ? ['workspace'] : [],
       ...Array.from({ length: scenario.childSessions ?? 0 }, (_, index) => `session.${index + 1}.jsonl`),
     ].sort()
     expect((await readdir(scenarioDir(scenario))).sort()).toEqual(expected)
     for (const fixture of ['session.jsonl', ...childFixturePaths(scenario).map(path => basename(path))]) {
-      const content = await readFile(join(scenarioDir(scenario), fixture), 'utf8')
+      const content = await readFile(join(fixtureDir(scenario), fixture), 'utf8')
       expect(scrubRequestHeaders(content), `${scenario.name}/${fixture} carries request-header bulk`).toBe(content)
     }
   }

apps/web/tests/live-interactions.e2e.ts

@@ -28,14 +28,14 @@ import { connectFreshWorkspace, newEnglishPage, saveFailureShot } from './suppor
 const SNAPSHOT_DIR = fileURLToPath(new URL('./snapshots/live-interactions', import.meta.url))
 const FIXTURE = join(SNAPSHOT_DIR, 'session.jsonl')
 // One golden pins the stable mid-turn loading state; the other three capture
-// what the user is left looking at after cancel, after a non-retryable failure
-// (pins the FIXME(web-error-surface) gap as a reviewable artifact: NO error
-// copy in the tree), and after retry recovery.
+// what the user is left looking at after cancel, after a non-retryable failure,
+// and after retry recovery.
 const CANCEL_EXPECTED = join(SNAPSHOT_DIR, 'cancel.expected.md')
 const LOADING_EXPECTED = join(SNAPSHOT_DIR, 'loading.expected.md')
 const ERROR_EXPECTED = join(SNAPSHOT_DIR, 'error-auth.expected.md')
 const RETRY_EXPECTED = join(SNAPSHOT_DIR, 'retry.expected.md')
 const MODE = webSnapshotMode()
+const AUTH_PROVIDER_MESSAGE = 'Authentication Fails, Your api key: sk-preview-secret is invalid'
 
 // The recorded base: one text-only turn whose derived script the sidecars
 // patch. Kept deliberately tool-free so the derived script is exactly one
@@ -158,7 +158,7 @@ describe('web e2e: live-turn interactions (cancel / error / retry)', () => {
 
   it.skipIf(MODE === 'record')('surfaces a non-retryable AUTH failure without retrying', async () => {
     await launch(() => ({
-      patches: [{ at: 0, entry: { kind: 'throw', chunks: [], message: 'invalid api key', code: 'AUTH' } }],
+      patches: [{ at: 0, entry: { kind: 'throw', chunks: [], message: AUTH_PROVIDER_MESSAGE, code: 'AUTH' } }],
     }))
     onTestFailed(() => saveFailureShot(page, 'web-e2e-error-auth'))
     const { settled } = await sendPrompt()
@@ -166,28 +166,28 @@ describe('web e2e: live-turn interactions (cancel / error / retry)', () => {
     expect(turnEndReasons(sessionEvents).at(-1)).toBe('error')
     // AUTH is outside llm-retry's retryable set: no retry record.
     expect(sessionEvents.filter(e => e.type === 'llm/retry').length).toBe(0)
-    // Product gap found by this lane, pinned as-is: the client consumes no
-    // agent/error frames and a pre-chunk failure freezes no partial, so THIS
-    // failure renders no error copy anywhere — the user sees the send simply
-    // stop. FIXME(web-error-surface): assert visible error text here once the
-    // web UI grows an error rendering; until then the pinned contract is
-    // "no crash, composer recovers, turn logged as error".
     await expect.poll(() => page.locator('textarea').first().isEnabled(), { timeout: 10_000 }).toBe(true)
     expect(await page.locator('[data-streaming="true"]').count()).toBe(0)
-    // The blank workspace also has an enabled composer. Wait for the driven
-    // session's only visible message before capturing its no-error-copy state.
-    await expect.poll(() => page.getByText(PROMPT, { exact: true }).first().isVisible(), { timeout: 10_000 }).toBe(true)
-    // Golden of the same gap: the prompt bubble alone, no error copy in the
-    // tree — the diff that changes when web-error-surface lands.
+    const errorStatus = page.getByRole('status').filter({ hasText: 'This turn failed' })
+    await errorStatus.waitFor({ timeout: 10_000 })
+    expect(await errorStatus.textContent()).toContain('API key is invalid')
+    expect(await errorStatus.textContent()).toContain('AUTH')
+    expect(await page.locator('body').textContent()).not.toContain('sk-preview-secret')
     const snapshot = await captureStableAria(page, '[class*="centerCol"]', scaffold!.workspaceCwd)
     await compareOrRefreshGolden(ERROR_EXPECTED, snapshot, MODE)
+    await page.getByRole('tab', { name: 'Trajectory' }).click()
+    const requestMarker = page.locator('tr[data-request-only="true"]').last()
+      .getByRole('button', { name: /Request #/ })
+    await requestMarker.click()
+    await page.getByText('API key is invalid', { exact: true }).waitFor({ timeout: 10_000 })
+    expect(await page.locator('body').textContent()).not.toContain('sk-preview-secret')
     expect(tripwire.pageErrors).toEqual([])
     expect(tripwire.warnings).toEqual([])
   }, 120_000)
 
   it.skipIf(MODE === 'record')('keeps a terminal request marker inside the trajectory table', async () => {
     await launch(() => ({
-      patches: [{ at: 0, entry: { kind: 'throw', chunks: [], message: 'invalid api key', code: 'AUTH' } }],
+      patches: [{ at: 0, entry: { kind: 'throw', chunks: [], message: AUTH_PROVIDER_MESSAGE, code: 'AUTH' } }],
     }))
     const { settled } = await sendPrompt()
     await settled

apps/web/tests/shipped-composition.e2e.ts

@@ -0,0 +1,83 @@
+// Boots the shipped Web composition over the built dist this lane already uses
+// and asserts what that composition produces: the model-visible tool catalog
+// and the sandbox/approval knobs it ships with. No browser and no model call —
+// these are composition facts, and the browser scenarios in this lane cover the
+// surface itself.
+import { tmpdir } from 'node:os'
+import { afterEach, expect, it } from 'vitest'
+import { canonicalPath, writableRoots } from '@deepseek-ai/dsh-sandbox'
+// Empty type imports carry the tools/sandboxPolicy/approval Context merges.
+import type {} from '@deepseek-ai/dsh-tools'
+import type {} from '@deepseek-ai/dsh-sandbox-policy'
+import type {} from '@deepseek-ai/dsh-user-approval'
+import { launchWebScaffold, type WebScaffold } from './scaffold.ts'
+
+/**
+ * The catalog the shipped Web composition puts in front of the model, minus the
+ * ripgrep-dependent pair below. The absences are deliberate, not incidental
+ * gaps: the `cordis_*` toolset executes model-written JavaScript that no
+ * sandbox row confines, `web_fetch` chooses its own request target, and
+ * `mcp_*` servers spawn outside `ctx.bash`. The composition Agent Note owns the
+ * rationale and its sources.
+ */
+const EXPECTED_TOOLS = [
+  'ask_user_question',
+  'bash',
+  'create_goal',
+  'edit',
+  'exit_plan_mode',
+  'get_goal',
+  'ralph',
+  'read',
+  'session_event_read',
+  'session_event_search',
+  'session_event_trace',
+  'session_search',
+  'session_trace',
+  'skill',
+  'str_replace_editor',
+  'subagent',
+  'subagent_fork',
+  'task_kill',
+  'task_list',
+  'task_output',
+  'todo_write',
+  'update_goal',
+  'web_search',
+  'workflow',
+  'write',
+]
+
+/**
+ * `glob` and `grep` come from `dsh-tool-fs-search`, which probes `command -v rg`
+ * through the mounted bash executor at load and registers neither tool when
+ * ripgrep is absent. That is a host dependency, not a composition decision, so the
+ * pair is asserted separately — present together or absent together.
+ */
+const RIPGREP_TOOLS = ['glob', 'grep']
+
+let scaffold: WebScaffold | undefined
+
+afterEach(async () => {
+  await scaffold?.close()
+  scaffold = undefined
+})
+
+it('assembles the shipped Web catalog and keeps its access default', async () => {
+  scaffold = await launchWebScaffold()
+  const names = scaffold.ctx.tools.schemas().map(schema => schema.name).sort()
+  expect(names.filter(name => !RIPGREP_TOOLS.includes(name))).toEqual(EXPECTED_TOOLS)
+  expect([[], RIPGREP_TOOLS]).toContainEqual(names.filter(name => RIPGREP_TOOLS.includes(name)))
+  // `workspace-write` is not "the workspace and nothing else": the shared roots
+  // helper always admits the temp directories too. Pinning it against an
+  // explicit mode keeps the claim independent of this surface's default, and
+  // keeps a future boundary test from being run inside /tmp — where an
+  // "escape" write succeeds by design and reads as a sandbox failure.
+  expect(writableRoots(scaffold.ctx.sandboxPolicy.resolve({ mode: 'workspace-write' }))).toEqual(
+    expect.arrayContaining([canonicalPath('/tmp'), canonicalPath(tmpdir())]),
+  )
+  // The Web surface keeps its shipped access default; the base's confined one
+  // reaches the TUI. Pinning both keeps a base change from moving Web silently.
+  expect(scaffold.ctx.sandboxPolicy.defaultMode).toBe('danger-full-access')
+  expect(scaffold.ctx.approval.config.policy).toBe('never')
+}, 120_000)

apps/web/tests/snapshots/lifecycle-chrome/command-menu.expected.md

@@ -1,6 +1,7 @@
 - listbox "Trigger suggestions":
   - text: Commands
-  - option "goal set or view the goal for a long-running task" [selected]
+  - option "compact Compact older conversation history" [selected]
+  - option "goal set or view the goal for a long-running task"
   - option "permission Switch the permission preset (sandbox mode + approval policy)"
   - option "plan Enter or leave plan mode"
   - option "model Select the model for this conversation"

apps/web/tests/snapshots/live-interactions/error-auth.expected.md

@@ -13,6 +13,9 @@
   - img
   - img
   - text: Context injection
+- status:
+  - text: This turn failedAPI key is invalid
+  - code: AUTH
 - textbox "Message the agent"
 - button "Commands":
   - img

apps/web/tests/startup-auto-selection.e2e.ts

@@ -0,0 +1,119 @@
+// Web e2e scenario: startup auto-selection keeps the hero on screen.
+//
+// A page load with a workspace already registered runs
+// `WorkspacesService.startInitialSelection`: it connects the most recent
+// workspace and opens its blank session. `openState` flips to `loading` the
+// moment `open()` lands, which used to drive `data-phase=settling` on the
+// conversation root — `visibility:hidden` over the composer seat and the
+// header for the whole `session.history` round-trip, so the center column went
+// blank and repainted, reading as a full-page refresh on every launch.
+//
+// The unit spec pins the phase condition over hand-built stores. What only the
+// assembled application can show is that the path a user actually takes
+// reaches it: the real selection service, the real client session opening over
+// the real /api transport, and a real browser deciding what is painted.
+//
+// The round-trip against a loopback host is far too fast to observe, so this
+// scenario HOLDS the `session.history` response open at the browser's network
+// boundary and asserts the visible frame while it is in flight. That gate is
+// what makes the assertions non-vacuous: with the exemption reverted the held
+// window is exactly when `settling` is painted and the composer is hidden.
+//
+// Zero model calls: registering a workspace and opening its blank session are
+// host RPCs with no model involvement. A stray stream would fail loud with
+// NO_ADAPTER.
+import type { Browser, Page } from 'playwright'
+import { chromium } from 'playwright'
+import { afterAll, beforeAll, describe, expect, it, onTestFailed } from 'vitest'
+import { acknowledgeReloadConnectionLoss, launchWebScaffold, watchConsole, type WebScaffold } from './scaffold.ts'
+import { connectFreshWorkspace, newEnglishPage, saveFailureShot } from './support.ts'
+
+/** Wire path of the history round-trip the conversation root waits out (POST /api/session.history). */
+const HISTORY_ROUTE = '**/api/session.history'
+
+/**
+ * The conversation root's own phase attribute. `div` disambiguates it from the
+ * composer textarea, which carries an unrelated `data-phase` of its own.
+ */
+const ROOT_PHASE = 'div[data-phase]'
+
+/** Every distinct `data-phase` the conversation root shows, in order, across one page load. */
+function recordedPhases(page: Page): Promise<string[]> {
+  return page.evaluate(() => (window as unknown as { __conversationPhases: string[] }).__conversationPhases)
+}
+
+describe('web e2e: startup auto-selection', () => {
+  let scaffold: WebScaffold
+  let browser: Browser
+  let page: Page
+  let tripwire: ReturnType<typeof watchConsole>
+
+  beforeAll(async () => {
+    scaffold = await launchWebScaffold({})
+    browser = await chromium.launch()
+    page = await newEnglishPage(browser)
+    tripwire = watchConsole(page)
+    await page.goto(scaffold.baseUrl, { waitUntil: 'load' })
+    await page.waitForSelector('[class*="frame"]', { timeout: 30_000 })
+    // A registered workspace is the precondition for auto-selection: the first
+    // load has nothing to select, so the reload below is the path under test.
+    await connectFreshWorkspace(page, scaffold.workspaceCwd, 'startup-auto-selection')
+  }, 180_000)
+
+  afterAll(async () => {
+    await browser?.close()
+    await scaffold?.close()
+  })
+
+  it('keeps the hero and the composer on screen while the auto-selected blank session opens', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-startup-auto-selection'))
+    // Runs before any page script on the reload below, so the first phase the
+    // root ever renders is recorded, not just the ones after a listener attaches.
+    await page.addInitScript(() => {
+      const phases: string[] = []
+      ;(window as unknown as { __conversationPhases: string[] }).__conversationPhases = phases
+      setInterval(() => {
+        const phase = document.querySelector('div[data-phase]')?.getAttribute('data-phase')
+        if (phase === null || phase === undefined) return
+        if (phases[phases.length - 1] !== phase) phases.push(phase)
+      }, 8)
+    })
+
+    let releaseHistory = (): void => {}
+    const historyHeld = new Promise<void>((resolve) => { releaseHistory = resolve })
+    let historyRequested = (): void => {}
+    const historyInFlight = new Promise<void>((resolve) => { historyRequested = resolve })
+    let gated = false
+    await page.route(HISTORY_ROUTE, async (route) => {
+      // Only the auto-selection's own round-trip is held; later pages must not
+      // deadlock behind a gate this test has already released.
+      if (gated) { await route.continue(); return }
+      gated = true
+      historyRequested()
+      await historyHeld
+      await route.continue()
+    })
+
+    const warningsBefore = tripwire.warnings.length
+    await page.reload({ waitUntil: 'commit' })
+    await historyInFlight
+
+    // The frame a user sees while the session is still opening: hero phase, the
+    // hero title, and a composer that is actually painted (`settling` hides the
+    // seat with `visibility:hidden`, which Playwright reports as not visible).
+    await page.waitForSelector(ROOT_PHASE, { timeout: 15_000 })
+    expect(await page.locator(ROOT_PHASE).first().getAttribute('data-phase')).toBe('hero')
+    expect(await page.getByText("Let's start building").isVisible()).toBe(true)
+    expect(await page.locator('textarea').first().isVisible()).toBe(true)
+
+    releaseHistory()
+    await page.locator('textarea:enabled[placeholder="Describe what you want to build"]')
+      .waitFor({ timeout: 15_000 })
+    acknowledgeReloadConnectionLoss(tripwire, warningsBefore)
+
+    // Settling is not merely absent from the frame sampled above: the root
+    // never entered it at any point of the load.
+    expect(await recordedPhases(page)).toEqual(['hero'])
+    expect(tripwire.pageErrors).toEqual([])
+  }, 120_000)
+})

apps/web/tsconfig.json

@@ -48,7 +48,9 @@
     "tests/queue-actions.e2e.ts",
     "tests/skill-invocation-policy.e2e.ts",
     "tests/permission-policy-context.e2e.ts",
-    "tests/access-confirmation.e2e.ts"
+    "tests/access-confirmation.e2e.ts",
+    "tests/shipped-composition.e2e.ts",
+    "tests/startup-auto-selection.e2e.ts"
   ],
   "references": [
     {

docs/architecture.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 docs/architecture.md
-architecture.md: bcd94b1a6b361047d13fbdec41f3d0248b5f1821
-architecture.zh.md: 7bd21c617a6872f8db45d3f02ccddce56b43c0b6
+architecture.md: def0e8a3419b225c5f268de8415befcc5a937af3
+architecture.zh.md: ff1cb05ddfbf69af51faa401cfb306cb20952cae

docs/architecture.md

@@ -130,11 +130,11 @@ Adapter failures close their step before `agent/request-error` receives the exac
 
 Other failures use `agent/error`. Cancellation and disposal beat recovery. Before request-header commit, the turn signal cancels asynchronous model-capability preparation; undispatched tools get synthetic `tool/call`/`ABORTED_BEFORE_DISPATCH` pairs. Effective `cancel(cause)` emits its cause before queue clearing and abort; observers cannot veto; idle calls emit nothing. Durability records user or parent cancellation as `aborted`, teardown as `disposed`; teardown awaits quiescence. The cause affects reporting, not late result-context handling ([decision](../.agents/notes/implemented/architecture/2026-07-16-explicit-turn-cancellation.md)).
 
-Turn and step events are turn-enclosed; idle injected `user/message` events may sit between turns. Reload closes an interrupted tail with a synthetic turn end. After close, only `agent/error` reports failures. Each turn has one [TurnEndReason](core-data-structures/session.md#why-a-turn-ended-turnendreasonmap).
+Turn and step events are turn-enclosed. Idle `user/message` and standalone `compact/* { turn: null }` consume no turn; their lock-time markers may interleave with injection. Reload synthesizes interrupted turn ends; `session/end-seed` distinguishes stale compaction orphans from live locks. After close, only `agent/error` reports failures. Each turn has one [TurnEndReason](core-data-structures/session.md#why-a-turn-ended-turnendreasonmap).
 
 ### Agent Handles
 
-`ctx.agents` owns live agents and returns `AgentHandle { agent, dispose() }`. Plugins use full `send()` options or `followup()`, `steer()`, and `inject()` presets; `cancel()` and `whenIdle()` control lifecycle. One awaited disposer coordinates teardown ownership.
+`ctx.agents` owns agents, returning `AgentHandle { agent, dispose() }`. Plugins use `send()` or `followup()`, `steer()`, and `inject()` presets; [`reserveTurnAdmission()`](../packages/core/agent/README.md#agent-interface-typests) synchronously reserves idle for durable work without changing queued prompt identity. `cancel()` and `whenIdle()` control lifecycle. Awaited disposal owns teardown.
 
 ### Agent Scope
 
@@ -144,13 +144,13 @@ Each agent owns scoped `agent.ctx`; shared storage overlays its tool, prompt, an
 
 ### Session Log
 
-The session log is authoritative. `deriveMessages()` projects model history; `assistant/chunk` preserves replay and UI fidelity. Fork, resume, transcript rendering, telemetry, and persistence derive from it.
+The session log is authoritative. `deriveMessages()` projects model history; `assistant/chunk` preserves replay and UI fidelity. Fork, resume, transcripts, telemetry, and persistence derive from it.
 
-**Model-visible ⟺ durably recorded**: before `step/start`, the loop appends the full current runtime-context snapshot as a sourced `user/message`, then snapshots derived messages. Those messages, logged immutable attachment references, and the folded `request/header` reconstruct each request. The header marks adapter defaults so later proposals discard them and re-resolve the route without losing explicit settings. `dsh-agent-loop/invariant` asserts reconstructability through `ctx.invariants`, while attachment backends assert byte integrity ([decision](../.agents/notes/implemented/architecture/2026-07-05-reconstructable-requests.md)).
+**Model-visible ⟺ durably recorded**: before `step/start`, the loop appends the current runtime-context snapshot as a sourced `user/message`, then snapshots messages. Those messages, logged immutable attachment references, and folded `request/header` reconstruct each request. It marks adapter defaults so proposals discard them and re-resolve the route without losing explicit settings. `dsh-agent-loop/invariant` asserts reconstructability through `ctx.invariants`, while attachment backends assert byte integrity ([decision](../.agents/notes/implemented/architecture/2026-07-05-reconstructable-requests.md)).
 
 Plugins own durability. Backends eagerly drain synchronous `session/event` notifications. `session/flush` barriers precede each request and top-level tool dispatch, then follow `turn/end` before another queued turn or idle observation. `SessionPersistence` stores `SessionEvent` directly and metadata in `SessionHeader`; JSONL defaults to checksummed Zstandard, while SQLite shares the contract ([decision](../.agents/notes/implemented/bug-fix/2026-07-21-semantic-session-checkpoints.md)).
 
-Log-only events may sit between turns. Owners append through `Session`, flushing only for durability. `session/title` relies on eager persistence and lifecycle drains. Latest title wins with provenance; fallback and provider work never delays responses. Such records are fork boundaries, so forks inherit titles ([decision](../.agents/notes/implemented/feature/2026-07-21-log-backed-session-titles.md)).
+Between turns, owners append log-only events through `Session`, flushing only for durability. `session/title` needs eager persistence and lifecycle drains; manual compaction flushes its bracket before releasing admission. Title work never delays responses; latest wins with provenance. Title records are inherited fork boundaries ([decision](../.agents/notes/implemented/feature/2026-07-21-log-backed-session-titles.md)).
 
 ### Model Content
 

docs/architecture.zh.md

@@ -130,11 +130,11 @@ idle inject:
 
 其他故障使用 `agent/error`。取消和资源释放优先于恢复。在提交请求头之前，轮次信号会取消异步模型能力准备；尚未分派的工具会得到合成的 `tool/call`/`ABORTED_BEFORE_DISPATCH` 对。实际生效的 `cancel(cause)` 在清空队列和中止前发出原因；观察方不能否决；空闲调用不发事件。持久化层将用户或父级取消记录为 `aborted`，拆卸记录为 `disposed`；拆卸会等待完全停稳。原因只影响报告方式，不影响延迟完成的结果上下文处理（[决策](../.agents/notes/implemented/architecture/2026-07-16-explicit-turn-cancellation.md)）。
 
-轮次和步骤事件均位于轮次边界内；空闲时注入的 `user/message` 可以位于两个轮次之间。重新加载会用合成的轮次结束事件闭合中断尾部。关闭后仅由 `agent/error` 报告故障。每个轮次有一个 [TurnEndReason](core-data-structures/session.md#why-a-turn-ended-turnendreasonmap)。
+轮次和步骤事件均位于轮次边界内。空闲 `user/message` 与独立的 `compact/* { turn: null }` 不占用轮次；其锁定时刻标记可以与注入交错。重新加载会为中断的轮次合成结束事件；`session/end-seed` 区分陈旧的压缩遗留项与活跃锁。关闭后仅由 `agent/error` 报告故障。每个轮次有一个 [TurnEndReason](core-data-structures/session.md#why-a-turn-ended-turnendreasonmap)。
 
 ### Agent 句柄
 
-`ctx.agents` 拥有活跃 agent，并返回 `AgentHandle { agent, dispose() }`。插件使用全部 `send()` 选项，或 `followup()`、`steer()` 和 `inject()` 预设；`cancel()` 与 `whenIdle()` 控制生命周期。一个需等待完成的 disposer 协调拆卸归属。
+`ctx.agents` 拥有 agent，返回 `AgentHandle { agent, dispose() }`。插件使用 `send()`，或使用 `followup()`、`steer()` 和 `inject()` 预设；[`reserveTurnAdmission()`](../packages/core/agent/README.md#agent-interface-typests) 为持久工作同步预留空闲状态，同时不改变排队提示词身份。`cancel()` 与 `whenIdle()` 控制生命周期。需等待完成的资源释放负责拆卸。
 
 ### Agent 作用域
 
@@ -144,13 +144,13 @@ idle inject:
 
 ### 会话日志
 
-会话日志是权威依据。`deriveMessages()` 投影出模型历史；`assistant/chunk` 保证回放和 UI 保真。fork、恢复、transcript（文本记录）渲染、遥测和持久化均派生自该日志。
+会话日志是权威依据。`deriveMessages()` 投影出模型历史；`assistant/chunk` 保证回放和 UI 保真。fork、恢复、transcript（文本记录）、遥测和持久化均派生自该日志。
 
-**模型可见 ⟺ 已持久记录**：在 `step/start` 之前，循环会将完整的当前运行时上下文快照作为一条带来源的 `user/message` 追加，随后对派生消息制作快照。这些消息、不可变附件的日志引用与折叠后的 `request/header` 共同重建每个请求。该 header 会标记适配器默认值，使后续提议丢弃这些值并重新解析路由，同时不丢失显式设置。`dsh-agent-loop/invariant` 通过 `ctx.invariants` 断言可重建性，附件后端则断言附件字节完整性（[决策](../.agents/notes/implemented/architecture/2026-07-05-reconstructable-requests.md)）。
+**模型可见 ⟺ 已持久记录**：在 `step/start` 之前，循环会将当前运行时上下文快照作为一条带来源的 `user/message` 追加，随后对消息制作快照。这些消息、不可变附件的日志引用与折叠后的 `request/header` 共同重建每个请求。它会标记适配器默认值，使提议丢弃这些值并重新解析路由，同时不丢失显式设置。`dsh-agent-loop/invariant` 通过 `ctx.invariants` 断言可重建性，附件后端则断言附件字节完整性（[决策](../.agents/notes/implemented/architecture/2026-07-05-reconstructable-requests.md)）。
 
 插件负责持久性。后端会尽快排空同步的 `session/event` 通知。`session/flush` 屏障位于每次请求与顶层工具分发之前，并在 `turn/end` 之后、处理另一个已排队轮次或观察到空闲状态之前执行。`SessionPersistence` 直接存储 `SessionEvent`，并将元数据存入 `SessionHeader`；JSONL 默认采用带校验和的 Zstandard，SQLite 遵循同一契约（[决策](../.agents/notes/implemented/bug-fix/2026-07-21-semantic-session-checkpoints.md)）。
 
-纯日志事件可以位于轮次之间。事件所有方通过 `Session` 追加，仅为持久性而刷写。`session/title` 依赖尽快持久化与生命周期排空。最新标题按后写覆盖并携带来源信息；回退与提供方工作绝不会延迟响应。这类记录可作为 fork 边界，因此 fork 会继承标题（[决策](../.agents/notes/implemented/feature/2026-07-21-log-backed-session-titles.md)）。
+在轮次之间，事件所有方通过 `Session` 追加纯日志事件，仅为持久性而刷写。`session/title` 需要尽快持久化与生命周期排空；手动压缩会在释放轮次接纳预留前 flush 其标记对。标题工作绝不延迟响应；最新标题按后写覆盖并携带来源信息。标题记录是可继承的 fork 边界（[决策](../.agents/notes/implemented/feature/2026-07-21-log-backed-session-titles.md)）。
 
 ### 模型内容
 

docs/config-catalog.md

@@ -375,7 +375,7 @@ Source: [`packages/code-runtime/code-runtime-worker/src/index.ts:25`](../package
 
 ## `@deepseek-ai/dsh-compact-basic`
 
-Requires: `llm` · `tokenMeter`
+Requires: `llm` · `tokenMeter` · `sessions`
 
 ```ts config-catalog
 /** Basic compaction configuration with an optional exact-target policy table. */
@@ -2334,6 +2334,7 @@ These load from a `cordis.yml` entry with no `config:` block; they declare no co
 - `@deepseek-ai/dsh-client-ui-theme` ([`packages/client/ui-theme/src/index.ts`](../packages/client/ui-theme/src/index.ts))
 - `@deepseek-ai/dsh-client-ui-trajectory` ([`packages/client/ui-trajectory/src/index.ts`](../packages/client/ui-trajectory/src/index.ts))
 - `@deepseek-ai/dsh-client-ui-workspace` ([`packages/client/ui-workspace/src/index.ts`](../packages/client/ui-workspace/src/index.ts))
+- `@deepseek-ai/dsh-command-compact` — requires `commands` · `compact` ([`packages/compact/command-compact/src/index.ts`](../packages/compact/command-compact/src/index.ts))
 - `@deepseek-ai/dsh-command-goal` — requires `commands` · `goals` ([`packages/goal/command-goal/src/index.ts`](../packages/goal/command-goal/src/index.ts))
 - `@deepseek-ai/dsh-commands` ([`packages/ui/commands/src/index.ts`](../packages/ui/commands/src/index.ts))
 - `@deepseek-ai/dsh-fs-policy` ([`packages/fs/fs-policy/src/index.ts`](../packages/fs/fs-policy/src/index.ts))

docs/cordis-catalog/events.md

@@ -32,7 +32,7 @@ Effective broad cancellation was requested, before queued/outbox work is cleared
 
 Types: [Agent](../core-data-structures/core.md) · [AgentCancelCause](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:319`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:333`](../../packages/core/agent/src/types.ts)
 
 ### `agent/created` — emit
 
@@ -54,7 +54,7 @@ A fully configured agent and live session were published. Setup is composition-o
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:250`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:264`](../../packages/core/agent/src/types.ts)
 
 ### `agent/disposed` — emit
 
@@ -74,7 +74,7 @@ An agent left the registry; AgentLoop emits this after driver quiescence and sco
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:259`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:273`](../../packages/core/agent/src/types.ts)
 
 ### `agent/error` — emit
 
@@ -96,7 +96,7 @@ A step or turn errored. The machine reports a failure here (plus the logger) eve
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:433`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:447`](../../packages/core/agent/src/types.ts)
 
 ### `agent/inbox/dequeue` — emit
 
@@ -117,7 +117,7 @@ The driver claimed one item out of the inbox: a queued item at a turn boundary,
 
 Types: [Agent](../core-data-structures/core.md) · [InboxItem](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:297`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:311`](../../packages/core/agent/src/types.ts)
 
 ### `agent/inbox/discard` — emit
 
@@ -140,7 +140,7 @@ Pending inbox items were dropped without delivering them, so every enqueue occur
 
 Types: [Agent](../core-data-structures/core.md) · [InboxItem](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:309`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:323`](../../packages/core/agent/src/types.ts)
 
 ### `agent/inbox/enqueue` — emit
 
@@ -161,7 +161,7 @@ An item entered the queued or steering inbox. `placement` is the acceptance-time
 
 Types: [Agent](../core-data-structures/core.md) · [InboxItem](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:278`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:292`](../../packages/core/agent/src/types.ts)
 
 ### `agent/inbox/update` — emit
 
@@ -181,7 +181,7 @@ A still-pending queued item changed content. The item id, placement, and positio
 
 Types: [Agent](../core-data-structures/core.md) · [InboxItem](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:287`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:301`](../../packages/core/agent/src/types.ts)
 
 ### `agent/prompt-submit` — waterfall
 
@@ -204,7 +204,7 @@ Allow, rewrite, or block one claimed prompt before it becomes a user message or
 
 Types: [Agent](../core-data-structures/core.md) · [PromptDecision](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md) · [UserMessage](../core-data-structures/session.md)
 
-Source: [`packages/core/agent/src/types.ts:346`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:360`](../../packages/core/agent/src/types.ts)
 
 ### `agent/request` — waterfall
 
@@ -228,7 +228,7 @@ Replace the frozen call configuration. `await next()` yields the config the mach
 
 Types: [Agent](../core-data-structures/core.md) · [LlmCallConfig](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:372`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:386`](../../packages/core/agent/src/types.ts)
 
 ### `agent/request-error` — waterfall
 
@@ -258,7 +258,7 @@ Handle a model-request failure after its failed step has closed but before the f
 
 Types: [Agent](../core-data-structures/core.md) · [LlmFailure](../core-data-structures/llm-streaming.md) · [RequestError](../core-data-structures/core.md) · [RequestErrorAction](../core-data-structures/core.md) · [ResolvedRetryPolicy](../core-data-structures/llm-streaming.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:391`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:405`](../../packages/core/agent/src/types.ts)
 
 ### `agent/session-start` — emit
 
@@ -280,7 +280,7 @@ The session lifecycle began, once before the first turn. Use `agent.inject()` to
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md) · [SessionStartSource](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:332`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:346`](../../packages/core/agent/src/types.ts)
 
 ### `agent/settled` — emit
 
@@ -305,7 +305,7 @@ One drain chain reached its terminal turn: that turn's `turn/end` is already com
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md) · [SettleReason](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:420`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:434`](../../packages/core/agent/src/types.ts)
 
 ### `agent/status` — emit
 
@@ -325,7 +325,7 @@ Agent status changed (`idle` ⇄ `running`). `send()` does not enter `running` s
 
 Types: [Agent](../core-data-structures/core.md) · [AgentStatus](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:268`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:282`](../../packages/core/agent/src/types.ts)
 
 ### `agent/step` — serial
 
@@ -349,7 +349,7 @@ Awaited serial checkpoint before EVERY request of a turn is built (the first as
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:359`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:373`](../../packages/core/agent/src/types.ts)
 
 ### `agent/turn-stopping` — serial
 
@@ -375,7 +375,7 @@ The turn is about to close: the model owes no response (no live tool calls, no f
 
 Types: [Agent](../core-data-structures/core.md) · [Scoped](../core-data-structures/scope.md)
 
-Source: [`packages/core/agent/src/types.ts:406`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:420`](../../packages/core/agent/src/types.ts)
 
 ## `agent-loop/*`
 

docs/cordis-catalog/services.md

@@ -495,6 +495,26 @@ Abstract compaction service. Implementations own trigger policy, retention, and
  */
 abstract compactIfNeeded( agent: CompactAgentContext, trigger: CompactionTrigger, signal: AbortSignal, ): Promise<CompactionResult | null>
 
+/**
+ * Explicitly compact useful history even below automatic pressure thresholds.
+ * Implementations reserve idle turn admission synchronously before any
+ * asynchronous work, select a useful range without writing on a no-op, then
+ * append a standalone `compact/start` before summarization. That durable
+ * marker is the compaction lock until one `compact/end` attempt. Later waking
+ * prompts remain accepted in FIFO order and start only after the optional
+ * durability checkpoint and admission release. Context injected while the
+ * summary runs may sit between the marker pair; only the selected span must
+ * remain stable.
+ *
+ * @param agent - idle agent whose durable history should be compacted.
+ * @param signal - command-owned cancellation forwarded to summarization.
+ * @returns the compaction result, or `null` when no safe useful range exists.
+ * @throws {@link ManualCompactionError} for expected busy, changed-span,
+ * summarization/shrink, commit-stage, or persistence failures, and the exact
+ * abort reason when cancelled. Failed attempts remain visible in the log.
+ */
+abstract compactNow( agent: ManualCompactAgentContext, signal: AbortSignal, ): Promise<CompactionResult | null>
+
 /**
  * Forcibly compact a range of surface nodes into a single summary node.
  * `start` and `end` name an inclusive span by surface position, not numeric seq
@@ -518,7 +538,7 @@ abstract compactRegion( start: number, end: number, agent: CompactAgentContext,
 
 Types: [CompactionResult](../core-data-structures/compaction.md) · [CompactionTrigger](../core-data-structures/compaction.md)
 
-Source: [`packages/compact/compact/src/index.ts:45`](../../packages/compact/compact/src/index.ts)
+Source: [`packages/compact/compact/src/index.ts:80`](../../packages/compact/compact/src/index.ts)
 
 ## `ctx.credentials` — `Credentials` (abstract seam)
 
@@ -2378,7 +2398,7 @@ The concrete provider retains pi-tui, focus, and terminal lifecycle state. Plugi
 abstract openOverlay(request: TuiOverlayRequest): TuiOverlaySession
 ```
 
-Source: [`packages/ui/tui/src/index.ts:241`](../../packages/ui/tui/src/index.ts)
+Source: [`packages/ui/tui/src/index.ts:242`](../../packages/ui/tui/src/index.ts)
 
 ## `ctx.typert` — `TypertRegistry`
 

docs/core-data-structures/compaction.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 docs/core-data-structures/compaction.md
-compaction.md: 3ae4d7e50452b549654b7a7162141a4be505d784
-compaction.zh.md: 448c3aaf298b65ebe88565190c5b3978b975f5f3
+compaction.md: 070adc65fb25b2536c88701c375cc0b2a5308559
+compaction.zh.md: 9167882f63b2931ba3ce49697e0c87164394af89

docs/core-data-structures/compaction.md

@@ -2,7 +2,7 @@
 
 English | [中文](compaction.zh.md)
 
-The compaction seam — a [capability seam](../../.agents/notes/implemented/architecture/2026-06-13-capability-seams.md) split like bash: interface ([dsh-compact](../../packages/compact/compact), `ctx.compact`), implementation (a backend such as [dsh-compact-basic](../../packages/compact/compact-basic)), and consumer (a `/compact` tool, deferred). Compaction is **one optional capability**, not part of the agent-loop spine — so its vocabulary lives here, not in [core.md](core.md). A tokenizer- or template-based backend is a sibling package implementing the same interface. Unlike bash, the interface necessarily depends on `dsh-session` and `dsh-llm`: its verbs act on an agent-owned `Session`, and its durable summary event uses the `ContentBlock` vocabulary (see the [compaction capability-seam Agent Note](../../.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.md)).
+The compaction seam — a [capability seam](../../.agents/notes/implemented/architecture/2026-06-13-capability-seams.md) split like bash: interface ([dsh-compact](../../packages/compact/compact), `ctx.compact`), implementation (a backend such as [dsh-compact-basic](../../packages/compact/compact-basic)), and human consumer ([dsh-command-compact](../../packages/compact/command-compact)). Compaction is **one optional capability**, not part of the agent-loop spine — so its vocabulary lives here, not in [core.md](core.md). A tokenizer- or template-based backend is a sibling package implementing the same interface. Unlike bash, the interface necessarily depends on `dsh-session` and `dsh-llm`: its verbs act on an agent-owned `Session`, and its durable summary event uses the `ContentBlock` vocabulary (see the [compaction capability-seam Agent Note](../../.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.md)).
 
 Source: [`packages/compact/compact/src/types.ts`](../../packages/compact/compact/src/types.ts)
 
@@ -12,12 +12,14 @@ Compaction extends [`SessionEventMap`](session.md) with three event types via de
 
 | Event | Payload | Role |
 |---|---|---|
-| `compact/start` | `{ turn }` | acquires the log-recorded lock |
+| `compact/start` | `{ turn }` | acquires the log-recorded lock; a number identifies the open automatic turn, while `null` identifies a standalone manual attempt |
 | `compact/summary` | `{ summary, rawOutput?, shadowedRange, shadowedSeqs, shadowedTokenCount, provider, model, maxTokens?, usage? }` | provenance: the safe summary projection, optional complete provider output and usage, the shadowed surface-boundary pair (`start`/`end` seqs — a position span, not a numeric interval), the shadowed seqs in surface order, the estimated token count, and the summarize call's envelope (`provider`, `model`, plus its generation cap when one applied) — logged so the one-shot request is reconstructable from log + code (the reconstructability Agent Note) |
-| `compact/end` | `{ turn, error? }` | releases the lock (`error` set when summarization threw) |
+| `compact/end` | `{ turn, error? }` | releases the lock with the same numeric-or-null owner (`error` records an unsuccessful attempt) |
 
 The lock brackets the **whole** operation: `compact/start` is appended first, then summarization, the `compact/summary` provenance record, and the `user/message` replacement all land, and only then `compact/end`. Releasing the lock last turns a crash mid-operation into a detectable orphaned lock (a `compact/start` with no matching `compact/end`) rather than a `compact/end` that falsely claims compaction finished.
 
+The markers are lock time points, not an exclusive container. An unrelated idle injection can appear between a standalone manual start and end while summarization is pending. The manual path revalidates only its selected positional span, so that injected context survives after the replacement checkpoint. A live unmatched start blocks every entry point; an unmatched start before a newer `session/end-seed` is stale evidence from a prior lifecycle and is ignored.
+
 These variants are merged inside a `declare module '@deepseek-ai/dsh-session'` block, so — unlike the top-level types on the other sub-pages — they are not pasted as a drift-checked ` ```ts type-equiv ` block (the `verify-type-equiv` extractor matches only top-level declarations by name). The payload table above is the catalog entry; follow the source link for the authoritative shapes.
 
 ## `CompactionResult`
@@ -60,7 +62,16 @@ Automatic callers state why policy is running; implementations may treat confirm
 type CompactionTrigger = 'pressure' | 'context-overflow'
 ```
 
-`CompactService` exposes `compactIfNeeded(agent, trigger, signal)` for automatic `pressure` or `context-overflow` policy, returning `null` when no safe work exists, and `compactRegion(...)` for an explicit inclusive surface range. Every backend marks its replacement `user/message` with `COMPACT_CHECKPOINT_SOURCE`; client and wire consumers import that value and `isCompactCheckpointSource()` from the cordis-free `@deepseek-ai/dsh-compact/checkpoint` subpath, while the package root re-exports both for host consumers. Implementations must forward the supplied signal to summarization. The seam owns no pricing API: the singleton [`ctx.tokenMeter`](token-meter.md) directly owns estimation and replay, while `dsh-compact-basic` owns retention, event sequencing, routed summarization calls, and their configuration.
+`CompactService` exposes `compactIfNeeded(agent, trigger, signal)` for automatic `pressure` or `context-overflow` policy, `compactNow(agent, signal)` for one useful idle-session reduction even below pressure, and `compactRegion(...)` for an explicit inclusive surface range. `compactNow()` synchronously reserves the agent's next-turn admission, returns `null` without writing when no useful range exists, records a standalone `turn: null` bracket before summarization, flushes a closed attempt, and then releases admission so ordinary queued prompts derive from the new surface. Every backend marks its replacement `user/message` with `COMPACT_CHECKPOINT_SOURCE`; client and wire consumers import that value and `isCompactCheckpointSource()` from the cordis-free `@deepseek-ai/dsh-compact/checkpoint` subpath, while the package root re-exports both for host consumers. The predicate keeps checkpoint recognition independent of any one backend. Implementations must forward the supplied signal to summarization. The seam owns no pricing API: the singleton [`ctx.tokenMeter`](token-meter.md) directly owns estimation and replay, while `dsh-compact-basic` owns retention, event sequencing, routed summarization calls, and their configuration.
+
+Expected manual failures use `ManualCompactionErrorCode`:
+
+```ts type-equiv
+/** Expected failure classes for an explicit idle-session compaction request. */
+type ManualCompactionErrorCode = 'busy' | 'changed' | 'summary' | 'commit' | 'persistence'
+```
+
+`changed` and `summary` leave the conversation surface unchanged but still close and persist the failed attempt in the log. `commit` may follow partial mutation; `persistence` means the in-memory bracket closed but its flush failed. Cancellation remains separate and throws the exact abort reason after required cleanup.
 
 Pressure compaction runs at serial `agent/step` before request derivation. Once pressure or canonical overflow qualifies, compact-basic invokes optional [`ctx.toolResultPrune`](../../packages/compact/compact-tool-result-prune/README.md) before range selection, remeasures through `ctx.tokenMeter`, and can advance the surface without a summary. Failed-request recovery runs through `agent/request-error` after the failed step closes and returns a retry action only when the surface replacement generation advances, even if later summary work throws after pruning; cancellation still wins. Region boundaries preserve tool-call/result pairing but not whole turns, allowing early closed steps of one oversized turn to compact. `dsh-compact-basic` owns thresholds, retained-tail policy, overflow caps, and failure handling.
 

docs/core-data-structures/compaction.zh.md

@@ -2,7 +2,7 @@
 
 [English](compaction.md) | 中文
 
-压缩 seam 是一个[能力 seam](../../.agents/notes/implemented/architecture/2026-06-13-capability-seams.md)，与 bash 一样分为接口（[dsh-compact](../../packages/compact/compact)，`ctx.compact`）、实现（例如 [dsh-compact-basic](../../packages/compact/compact-basic) 后端）和消费方（延期实现的 `/compact` 工具）。压缩是**一项可选能力**，不属于 agent loop（智能体循环）主干，因此其词汇定义在此而非 [core.md](core.md) 中。基于 tokenizer 或模板的后端是实现同一接口的兄弟包（package）。与 bash 不同，该接口必然依赖 `dsh-session` 和 `dsh-llm`：其动词作用于 agent 所有的 `Session`，而其持久摘要事件使用 `ContentBlock` 词汇（见[压缩能力 seam Agent Note（agent 决策记录）](../../.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.md)）。
+压缩 seam 是一个[能力 seam](../../.agents/notes/implemented/architecture/2026-06-13-capability-seams.md)，与 bash 一样分为接口（[dsh-compact](../../packages/compact/compact)，`ctx.compact`）、实现（例如 [dsh-compact-basic](../../packages/compact/compact-basic) 后端）和面向用户的消费方（[dsh-command-compact](../../packages/compact/command-compact)）。压缩是**一项可选能力**，不属于 agent loop（智能体循环）主干，因此其词汇定义在此而非 [core.md](core.md) 中。基于 tokenizer 或模板的后端是实现同一接口的兄弟包（package）。与 bash 不同，该接口必然依赖 `dsh-session` 和 `dsh-llm`：其动词作用于 agent 所有的 `Session`，而其持久摘要事件使用 `ContentBlock` 词汇（见[压缩能力 seam Agent Note（agent 决策记录）](../../.agents/notes/implemented/feature/2026-06-18-compaction-capability-seam.md)）。
 
 源码：[`packages/compact/compact/src/types.ts`](../../packages/compact/compact/src/types.ts)
 
@@ -12,12 +12,14 @@
 
 | 事件 | 载荷 | 作用 |
 |---|---|---|
-| `compact/start` | `{ turn }` | 获取日志记录的锁 |
+| `compact/start` | `{ turn }` | 获取日志记录的锁；数字标识打开的自动轮次，`null` 标识独立手动尝试 |
 | `compact/summary` | `{ summary, rawOutput?, shadowedRange, shadowedSeqs, shadowedTokenCount, provider, model, maxTokens?, usage? }` | provenance：安全摘要投影、可选的完整 provider 输出与 usage、被遮蔽的 surface 边界对（`start`/`end` seq——位置跨度，而非数值区间）、按 surface 顺序排列的被遮蔽 seq、估算 token 数，以及摘要调用的 envelope（`provider`、`model`，若有生成上限则还包括该上限）——写入日志后，该一次性请求可由日志 + 代码重建（见可重建性 Agent Note） |
-| `compact/end` | `{ turn, error? }` | 释放锁（摘要调用抛出异常时设置 `error`） |
+| `compact/end` | `{ turn, error? }` | 使用相同的数字或 `null` 归属值释放锁（`error` 记录失败尝试） |
 
 锁括住**整个**操作：先追加 `compact/start`，然后执行摘要生成、写入 `compact/summary` 来源记录与 `user/message` 替换，最后才追加 `compact/end`。最后释放锁意味着操作中途崩溃会表现为可检测的遗留锁（有 `compact/start` 而无匹配的 `compact/end`），而非一个虚假声称压缩已完成的 `compact/end`。
 
+这些标记表示锁的时间点，而不是排他的容器。摘要等待期间，不相关的空闲注入可以出现在独立的手动 start 与 end 之间。手动路径只重新验证所选位置 span，因此替换检查点之后仍保留该注入上下文。活动的未匹配 start 会阻塞所有入口点；较新 `session/end-seed` 之前的未匹配 start 是先前生命周期留下的陈旧证据，会被忽略。
+
 这些变体在 `declare module '@deepseek-ai/dsh-session'` 块内合并，因此——与其他子页面上的顶层类型不同——它们不以漂移检查的 ` ```ts type-equiv ` 块粘贴（`verify-type-equiv` 提取器只按名称匹配顶层声明）。上方的载荷表即为目录条目；权威形状请循源码链接查看。
 
 ## `CompactionResult`
@@ -60,7 +62,16 @@ interface CompactionResult {
 type CompactionTrigger = 'pressure' | 'context-overflow'
 ```
 
-`CompactService` 暴露 `compactIfNeeded(agent, trigger, signal)` 以执行自动 `pressure` 或 `context-overflow` 策略；没有可安全执行的工作时返回 `null`。它还针对显式、两端均包含的 surface 范围暴露 `compactRegion(...)`。每个后端都使用 `COMPACT_CHECKPOINT_SOURCE` 标记其替换用的 `user/message`；client 与 wire 消费方从无 cordis 的 `@deepseek-ai/dsh-compact/checkpoint` 子路径导入该值和 `isCompactCheckpointSource()`，包根则为 host 消费方重新导出两者。实现必须把传入的 signal 转发给摘要流程。该 seam 不拥有计价 API：单例 [`ctx.tokenMeter`](token-meter.md) 直接拥有估算与回放，而 `dsh-compact-basic` 拥有保留策略、事件排序、按路由执行的摘要调用及其配置。
+`CompactService` 暴露 `compactIfNeeded(agent, trigger, signal)` 以执行自动 `pressure` 或 `context-overflow` 策略，暴露 `compactNow(agent, signal)` 以便即使未达到压力也对空闲会话进行一次有效缩减，还针对显式、两端均包含的 surface 范围暴露 `compactRegion(...)`。`compactNow()` 会同步预留 agent 的下一轮次接纳；没有有效范围时返回 `null` 且不写入；在摘要前记录独立的 `turn: null` 标记对；flush 已闭合尝试；随后释放接纳预留，使普通排队提示词从新表层派生。每个后端都使用 `COMPACT_CHECKPOINT_SOURCE` 标记其替换用的 `user/message`；client 与 wire 消费方从无 cordis 的 `@deepseek-ai/dsh-compact/checkpoint` 子路径导入该值和 `isCompactCheckpointSource()`，包根则为 host 消费方重新导出两者。该判定函数使检查点识别不依赖任一特定后端。实现必须把传入的 signal 转发给摘要流程。该 seam 不拥有计价 API：单例 [`ctx.tokenMeter`](token-meter.md) 直接拥有估算与回放，而 `dsh-compact-basic` 拥有保留策略、事件排序、按路由执行的摘要调用及其配置。
+
+预期的手动失败使用 `ManualCompactionErrorCode`：
+
+```ts type-equiv
+/** Expected failure classes for an explicit idle-session compaction request. */
+type ManualCompactionErrorCode = 'busy' | 'changed' | 'summary' | 'commit' | 'persistence'
+```
+
+`changed` 和 `summary` 保持会话表层不变，但仍会闭合失败尝试并将其持久化到日志。`commit` 可能发生在部分变更之后；`persistence` 表示内存中的标记对已闭合，但 flush 失败。取消独立于这些失败，并在完成必要清理后抛出原始 abort 原因。
 
 压力压缩在串行 `agent/step` 中运行，先于请求推导。一旦压力或规范化溢出满足条件，compact-basic 会在选择范围前调用可选的 [`ctx.toolResultPrune`](../../packages/compact/compact-tool-result-prune/README.md)，再通过 `ctx.tokenMeter` 重新测量，并且可以在不生成摘要的情况下推进 surface。失败请求的恢复在失败的步骤关闭后通过 `agent/request-error` 运行；仅当 surface replacement generation 前进时才返回重试动作，即便后续摘要工作在剪枝后抛异常亦如此；取消仍然优先。区域边界保持工具调用/结果配对，但不保持整个轮次，因此一个过大轮次中较早关闭的步骤可以被压缩。`dsh-compact-basic` 拥有阈值、保留尾部策略、溢出上限与失败处理。
 

docs/core-data-structures/core.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 docs/core-data-structures/core.md
-core.md: 19fc1045141e366f3baebceba1c0470d588cb912
-core.zh.md: 01e3dd48f3cfe0fb38f9248da2a0bf04c745d00e
+core.md: f88d7d6cd7537a5bba942c6c2f22a5268e930049
+core.zh.md: 3fcd0477364800406cda258dea833d536d31e4fa

docs/core-data-structures/core.md

@@ -618,6 +618,20 @@ interface Agent {
    */
   send(message: UserMessage, options: SendOptions): void
 
+  /**
+   * Reserve admission of the next ordinary turn while this agent is idle, so an
+   * operation can mutate durable history before any queued prompt derives a
+   * request from it. Already-accepted waking work has right of way, including a
+   * send whose wake is still a pending microtask. Later sends keep their
+   * ordinary placement, FIFO order, and `wakeup` facts, and
+   * {@link acceptsNextStep} stays `false`, so a waking `next-step` send becomes
+   * a queued follow-up rather than steering; cancellation and disposal may
+   * still discard them. {@link inject} is not withheld. {@link whenIdle} treats
+   * a live reservation as activity, while lifecycle teardown does not await it.
+   * @returns the idempotent release, or `undefined` when the agent is running, already reserved, or already committed to waking work.
+   */
+  reserveTurnAdmission(): (() => void) | undefined
+
   /**
    * Mutate one still-pending queued occurrence synchronously. Editing preserves
    * the message identity and queue position; removal publishes its terminal
@@ -675,7 +689,7 @@ interface Agent {
 }
 ```
 
-`AgentStatus` is `'idle' | 'running'`, and `SessionId` is branded. Disposal removes the agent from the registry and emits `agent/disposed`; it is not a terminal status value. `running` describes the driver-wide drain interval and may span consecutive queued turns; it does not prove a turn is still open. `acceptsNextStep` is the narrower routing predicate for callers that must choose between steering the current admission/turn and submitting a fresh admitted prompt. `AgentOptions` is merge-extensible: core declares `provider?`, `model?`, and `maxTokens?` (dispatch requires provider and model after `agent/request`). When present, `maxTokens` must be a positive safe integer and caps every conversation-model request; omission allows the exact-model adapter default to materialize before the request header, or otherwise leaves provider behavior unchanged. Persona belongs to `dsh-system-prompt`: an agent-scoped `deployment:persona` may shadow the global default.
+`AgentStatus` is `'idle' | 'running'`, and `SessionId` is branded. Disposal removes the agent from the registry and emits `agent/disposed`; it is not a terminal status value. `running` describes the driver-wide drain interval and may span consecutive queued turns; it does not prove a turn is still open. `acceptsNextStep` is the narrower routing predicate for callers that must choose between steering the current admission/turn and submitting a fresh admitted prompt. A live turn-admission reservation is quiescence-relevant without changing `status` or turning later queue entries into steering; its only authority is to defer the next driver claim until release. `AgentOptions` is merge-extensible: core declares `provider?`, `model?`, and `maxTokens?` (dispatch requires provider and model after `agent/request`). When present, `maxTokens` must be a positive safe integer and caps every conversation-model request; omission allows the exact-model adapter default to materialize before the request header, or otherwise leaves provider behavior unchanged. Persona belongs to `dsh-system-prompt`: an agent-scoped `deployment:persona` may shadow the global default.
 
 The cause is a TypeScript-enforced same-process input. An active `TurnCancellation` holder copies its discriminant into the runtime-only `AbortSignal.reason` and is retired before `turn/end` publication; the frozen `AbortSignal.reason` remains readable after that retirement. Only the loop reads the cause (`user`, `parent`, or lifecycle-only `disposed`) back off its own machine-private signal at settlement — there is no public reader, and a signal grants cooperating listeners no classification authority. Durable `turn/end` retains the coarse `{ kind: 'aborted' }` outcome; request provenance would require a separate durable event rather than overloading the terminal result.
 

docs/core-data-structures/core.zh.md

@@ -626,6 +626,20 @@ interface Agent {
    */
   send(message: UserMessage, options: SendOptions): void
 
+  /**
+   * Reserve admission of the next ordinary turn while this agent is idle, so an
+   * operation can mutate durable history before any queued prompt derives a
+   * request from it. Already-accepted waking work has right of way, including a
+   * send whose wake is still a pending microtask. Later sends keep their
+   * ordinary placement, FIFO order, and `wakeup` facts, and
+   * {@link acceptsNextStep} stays `false`, so a waking `next-step` send becomes
+   * a queued follow-up rather than steering; cancellation and disposal may
+   * still discard them. {@link inject} is not withheld. {@link whenIdle} treats
+   * a live reservation as activity, while lifecycle teardown does not await it.
+   * @returns the idempotent release, or `undefined` when the agent is running, already reserved, or already committed to waking work.
+   */
+  reserveTurnAdmission(): (() => void) | undefined
+
   /**
    * Mutate one still-pending queued occurrence synchronously. Editing preserves
    * the message identity and queue position; removal publishes its terminal
@@ -683,7 +697,7 @@ interface Agent {
 }
 ```
 
-`AgentStatus` 为 `'idle' | 'running'`，`SessionId` 是品牌类型。dispose（资源释放）会把 agent 从注册表移除并发出 `agent/disposed`；它不是一个终态 status 值。`running` 描述整个驱动器的排空区间，可能跨越连续的排队轮次；它不能证明某个轮次仍然打开。对于需要在把输入作为 steering 加入当前提示词准入／轮次，还是提交为一个新的待准入提示词之间做选择的调用方，`acceptsNextStep` 才是更窄且准确的路由判断条件。`AgentOptions` 可合并扩展：core 声明 `provider?`、`model?` 与 `maxTokens?`（在 `agent/request` 后，分发要求 provider 与 model 都存在）。提供 `maxTokens` 时，它必须是正安全整数，并限制每次对话模型请求的输出；省略时，系统会在写入请求 header 前填入确切模型的适配器默认值，否则提供方行为保持不变。Persona 归 `dsh-system-prompt` 所有：agent 作用域的 `deployment:persona` 可以遮蔽全局默认值。
+`AgentStatus` 为 `'idle' | 'running'`，`SessionId` 是品牌类型。dispose（资源释放）会把 agent 从注册表移除并发出 `agent/disposed`；它不是一个终态 status 值。`running` 描述整个驱动器的排空区间，可能跨越连续的排队轮次；它不能证明某个轮次仍然打开。对于需要在把输入作为 steering 加入当前提示词准入／轮次，还是提交为一个新的待准入提示词之间做选择的调用方，`acceptsNextStep` 才是更窄且准确的路由判断条件。活动的轮次接纳预留与完全停稳相关，但不会改变 `status`，也不会把之后的队列项变成 steering；它的唯一权限是将驱动器的下一次认领延迟到释放时。`AgentOptions` 可合并扩展：core 声明 `provider?`、`model?` 与 `maxTokens?`（在 `agent/request` 后，分发要求 provider 与 model 都存在）。提供 `maxTokens` 时，它必须是正安全整数，并限制每次对话模型请求的输出；省略时，系统会在写入请求 header 前填入确切模型的适配器默认值，否则提供方行为保持不变。Persona 归 `dsh-system-prompt` 所有：agent 作用域的 `deployment:persona` 可以遮蔽全局默认值。
 
 cause 是由 TypeScript 强制约束的同进程输入。活跃的 `TurnCancellation` 持有者会把其判别字段复制到仅运行时的 `AbortSignal.reason`，并在发布 `turn/end` 前退役；冻结后的 `AbortSignal.reason` 仍可读取。只有 loop 会在结算时从自己机器私有的 signal 上读回 cause（`user`、`parent` 或仅用于生命周期的 `disposed`）——不存在公开的读取器，signal 也不授予协作监听器任何分类权限。持久 `turn/end` 保留粗粒度 `{ kind: 'aborted' }` 结果；若需记录请求 provenance，应使用单独的持久事件，而不是让终态结果承担额外含义。
 

docs/event-producer-consumer.md

@@ -8,22 +8,22 @@ This matrix shows which packages dispatch each harness-owned event and which pac
 | Event | Mode | Declared in | Dispatchers | Listeners |
 | --- | --- | --- | --- | --- |
 | `agent-loop/config-start-failed` | `emit` | [`packages/core/agent-loop/src/index.ts:157`](../packages/core/agent-loop/src/index.ts) | [`agent-loop`](../packages/core/agent-loop) (`events.dispatch`) | [`tui`](../packages/ui/tui) |
-| `agent/cancel-requested` | `emit` | [`packages/core/agent/src/types.ts:319`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`goal-session`](../packages/goal/goal-session) |
-| `agent/created` | `emit` | [`packages/core/agent/src/types.ts:250`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`goal-session`](../packages/goal/goal-session), [`tui`](../packages/ui/tui) |
-| `agent/disposed` | `emit` | [`packages/core/agent/src/types.ts:259`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`agent-loop`](../packages/core/agent-loop), [`goal-session`](../packages/goal/goal-session), [`tui`](../packages/ui/tui) |
-| `agent/error` | `emit` | [`packages/core/agent/src/types.ts:433`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | `apiproxy`, [`goal-session`](../packages/goal/goal-session), [`session-telemetry`](../packages/telemetry/session-telemetry), [`tui`](../packages/ui/tui) |
-| `agent/inbox/dequeue` | `emit` | [`packages/core/agent/src/types.ts:297`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`tui`](../packages/ui/tui) |
-| `agent/inbox/discard` | `emit` | [`packages/core/agent/src/types.ts:309`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`tui`](../packages/ui/tui) |
-| `agent/inbox/enqueue` | `emit` | [`packages/core/agent/src/types.ts:278`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`goal-session`](../packages/goal/goal-session) |
-| `agent/inbox/update` | `emit` | [`packages/core/agent/src/types.ts:287`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | `apiproxy` |
-| `agent/prompt-submit` | `waterfall` | [`packages/core/agent/src/types.ts:346`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`goal-session`](../packages/goal/goal-session), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard), [`tui`](../packages/ui/tui) |
-| `agent/request` | `waterfall` | [`packages/core/agent/src/types.ts:372`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`agent`](../packages/core/agent) |
-| `agent/request-error` | `waterfall` | [`packages/core/agent/src/types.ts:391`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`compact-basic`](../packages/compact/compact-basic), [`llm-retry`](../packages/llm/llm-retry) |
-| `agent/session-start` | `emit` | [`packages/core/agent/src/types.ts:332`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`goal`](../packages/goal/goal), [`goal-session`](../packages/goal/goal-session), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`workspace-context`](../packages/context/workspace-context) |
-| `agent/settled` | `emit` | [`packages/core/agent/src/types.ts:420`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`compact-basic`](../packages/compact/compact-basic) |
-| `agent/status` | `emit` | [`packages/core/agent/src/types.ts:268`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`goal-session`](../packages/goal/goal-session), [`tui`](../packages/ui/tui) |
-| `agent/step` | `serial` | [`packages/core/agent/src/types.ts:359`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`compact-basic`](../packages/compact/compact-basic), [`plan-mode`](../packages/plan/plan-mode), [`session-checkpoint-policy`](../packages/session-persistence/session-checkpoint-policy), [`time-context`](../packages/context/time-context), [`tmux-context`](../packages/context/tmux-context), [`tool-skill`](../packages/skill/tool-skill), [`workspace-context`](../packages/context/workspace-context) |
-| `agent/turn-stopping` | `serial` | [`packages/core/agent/src/types.ts:406`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
+| `agent/cancel-requested` | `emit` | [`packages/core/agent/src/types.ts:333`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`goal-session`](../packages/goal/goal-session) |
+| `agent/created` | `emit` | [`packages/core/agent/src/types.ts:264`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`goal-session`](../packages/goal/goal-session), [`tui`](../packages/ui/tui) |
+| `agent/disposed` | `emit` | [`packages/core/agent/src/types.ts:273`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`agent-loop`](../packages/core/agent-loop), [`goal-session`](../packages/goal/goal-session), [`tui`](../packages/ui/tui) |
+| `agent/error` | `emit` | [`packages/core/agent/src/types.ts:447`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | `apiproxy`, [`goal-session`](../packages/goal/goal-session), [`session-telemetry`](../packages/telemetry/session-telemetry), [`tui`](../packages/ui/tui) |
+| `agent/inbox/dequeue` | `emit` | [`packages/core/agent/src/types.ts:311`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`tui`](../packages/ui/tui) |
+| `agent/inbox/discard` | `emit` | [`packages/core/agent/src/types.ts:323`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`tui`](../packages/ui/tui) |
+| `agent/inbox/enqueue` | `emit` | [`packages/core/agent/src/types.ts:292`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`goal-session`](../packages/goal/goal-session) |
+| `agent/inbox/update` | `emit` | [`packages/core/agent/src/types.ts:301`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | `apiproxy` |
+| `agent/prompt-submit` | `waterfall` | [`packages/core/agent/src/types.ts:360`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`goal-session`](../packages/goal/goal-session), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard), [`tui`](../packages/ui/tui) |
+| `agent/request` | `waterfall` | [`packages/core/agent/src/types.ts:386`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`agent`](../packages/core/agent) |
+| `agent/request-error` | `waterfall` | [`packages/core/agent/src/types.ts:405`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`compact-basic`](../packages/compact/compact-basic), [`llm-retry`](../packages/llm/llm-retry) |
+| `agent/session-start` | `emit` | [`packages/core/agent/src/types.ts:346`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`goal`](../packages/goal/goal), [`goal-session`](../packages/goal/goal-session), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`workspace-context`](../packages/context/workspace-context) |
+| `agent/settled` | `emit` | [`packages/core/agent/src/types.ts:434`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`compact-basic`](../packages/compact/compact-basic) |
+| `agent/status` | `emit` | [`packages/core/agent/src/types.ts:282`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emitAgentEvent`) | [`agent`](../packages/core/agent), `apiproxy`, [`goal-session`](../packages/goal/goal-session), [`tui`](../packages/ui/tui) |
+| `agent/step` | `serial` | [`packages/core/agent/src/types.ts:373`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`compact-basic`](../packages/compact/compact-basic), [`plan-mode`](../packages/plan/plan-mode), [`session-checkpoint-policy`](../packages/session-persistence/session-checkpoint-policy), [`time-context`](../packages/context/time-context), [`tmux-context`](../packages/context/tmux-context), [`tool-skill`](../packages/skill/tool-skill), [`workspace-context`](../packages/context/workspace-context) |
+| `agent/turn-stopping` | `serial` | [`packages/core/agent/src/types.ts:420`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
 | `approval/request` | `waterfall` | [`packages/ui/user-approval/src/index.ts:30`](../packages/ui/user-approval/src/index.ts) | [`user-approval`](../packages/ui/user-approval) (`waterfall`) | [`acp`](../packages/acp/acp), `apiproxy` |
 | `commands/change` | `emit` | [`packages/ui/commands/src/index.ts:154`](../packages/ui/commands/src/index.ts) | [`commands`](../packages/ui/commands) (`events.dispatch`) | `apiproxy`, [`tui`](../packages/ui/tui) |
 | `credentials/updated` | `emit` | [`packages/credentials/credentials/src/index.ts:67`](../packages/credentials/credentials/src/index.ts) | [`credentials`](../packages/credentials/credentials) (`events.dispatch`) | `apiproxy`, [`credentials`](../packages/credentials/credentials) |

docs/module-graph.md

@@ -57,6 +57,7 @@ flowchart TD
     pkg_tool_skill["tool-skill"]
   end
   subgraph group_compact["packages/compact"]
+    pkg_command_compact["command-compact"]
     pkg_compact["compact"]
     pkg_compact_basic["compact-basic"]
     pkg_compact_tool_result_prune["compact-tool-result-prune"]
@@ -628,6 +629,9 @@ flowchart TD
   pkg_fs_sandbox --> pkg_invariants
   pkg_fs_sandbox --> pkg_sandbox
   pkg_fs_sandbox --> pkg_sandbox_policy
+  pkg_command_compact --> pkg_commands
+  pkg_command_compact --> pkg_compact
+  pkg_command_compact --> pkg_invariants
   pkg_compact_basic --> pkg_agent
   pkg_compact_basic --> pkg_compact
   pkg_compact_basic --> pkg_compact_tool_result_prune
@@ -1172,6 +1176,7 @@ flowchart TD
 | [`goal-session`](../packages/goal/goal-session) | `goal` | [`agent`](../packages/core/agent), [`goal`](../packages/goal/goal), [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session) |
 | [`bash-sandbox`](../packages/bash/bash-sandbox) | `bash` | [`bash`](../packages/bash/bash), [`bash-local`](../packages/bash/bash-local), [`invariants`](../packages/support/invariants), [`sandbox`](../packages/sandbox/sandbox), [`sandbox-policy`](../packages/sandbox/sandbox-policy) |
 | [`fs-sandbox`](../packages/fs/fs-sandbox) | `fs` | [`fs`](../packages/fs/fs), [`fs-local`](../packages/fs/fs-local), [`invariants`](../packages/support/invariants), [`sandbox`](../packages/sandbox/sandbox), [`sandbox-policy`](../packages/sandbox/sandbox-policy) |
+| [`command-compact`](../packages/compact/command-compact) | `compact` | [`commands`](../packages/ui/commands), [`compact`](../packages/compact/compact), [`invariants`](../packages/support/invariants) |
 | [`compact-basic`](../packages/compact/compact-basic) | `compact` | [`agent`](../packages/core/agent), [`compact`](../packages/compact/compact), [`compact-tool-result-prune`](../packages/compact/compact-tool-result-prune), [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`token-meter`](../packages/llm/token-meter) |
 | [`session-query`](../packages/session-query/session-query) | `session-query` | [`brand`](../packages/util/brand), [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`session-title`](../packages/session-title/session-title) |
 | [`session-title-llm`](../packages/session-title/session-title-llm) | `session-title` | [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-title`](../packages/session-title/session-title), [`timeout`](../packages/util/timeout) |

docs/persistence-catalog.md

@@ -209,20 +209,27 @@ Source: [`packages/ui/commands/src/index.ts:132`](../packages/ui/commands/src/in
 #### `compact/end` — log-only
 
 ```ts persistence-catalog
-/** Marks the end of a compaction — log-only, releases the lock. `error` set if summarization failed. */
-'compact/end': { turn: number; error?: string }
+/**
+ * Marks the end of a compaction — log-only, releases the lock. Its owner
+ * matches `compact/start`; `error` records an unsuccessful attempt.
+ */
+'compact/end': { turn: number | null; error?: string }
 ```
 
-Source: [`packages/compact/compact/src/types.ts:44`](../packages/compact/compact/src/types.ts)
+Source: [`packages/compact/compact/src/types.ts:51`](../packages/compact/compact/src/types.ts)
 
 #### `compact/start` — log-only
 
 ```ts persistence-catalog
-/** Marks the start of a compaction — log-only, holds the lock until `compact/end`. */
-'compact/start': { turn: number }
+/**
+ * Marks the start of a compaction — log-only, holds the lock until
+ * `compact/end`. A numbered owner is strictly enclosed by that open turn;
+ * `null` identifies a standalone manual transaction between turns.
+ */
+'compact/start': { turn: number | null }
 ```
 
-Source: [`packages/compact/compact/src/types.ts:15`](../packages/compact/compact/src/types.ts)
+Source: [`packages/compact/compact/src/types.ts:19`](../packages/compact/compact/src/types.ts)
 
 #### `compact/summary` — log-only
 
@@ -258,7 +265,7 @@ Source: [`packages/compact/compact/src/types.ts:15`](../packages/compact/compact
 
 Types: [ContentBlock](core-data-structures/core.md) · [TokenUsage](core-data-structures/llm-streaming.md)
 
-Source: [`packages/compact/compact/src/types.ts:22`](../packages/compact/compact/src/types.ts)
+Source: [`packages/compact/compact/src/types.ts:26`](../packages/compact/compact/src/types.ts)
 
 ### `hook/*`
 

knip.json

@@ -627,7 +627,9 @@
         "tests/**/*.spec.ts",
         "tests/**/*.e2e.ts",
         "tests/**/*.snapshot.ts",
-        "tests/fixtures/tui-scripted-llm.ts"
+        "tests/fixtures/tui-scripted-llm.ts",
+        "tests/fixtures/composition-echo-llm.ts",
+        "tests/fixtures/composition-settled.ts"
       ],
       "project": [
         "src/**/*.ts",

packages/client/runtime/README.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 packages/client/runtime/README.md
-README.md: 2af844e812aca7692c9bdd9c4ca131481a4c0db0
-README.zh.md: ad18a97bde7ac1dabde2c4b67f25dd3eb60244be
+README.md: e862b024367137a937c1995137fb98e682838cf9
+README.zh.md: ae4ca7bc0be348feb719da10e26a17580338a170

packages/client/runtime/README.md

@@ -30,6 +30,10 @@ SlotsService gives the renderer separate bare observables for `useSessions` and
 
 Because the projection is log-ordered, the node array is seq-monotonic by construction: log-only `command/run` / `command/done` nodes splice in by seq, `Session` merges interrupted frozen nodes by their fractional seqs, and a window whose checkpoint cites a shadowed range outside it renders the marker with nothing logged. The marker's summary text comes from the checkpoint's `compact/summary` provenance; a window cut that left the provenance outside makes the row non-expandable rather than empty, and a later page that supplies it resolves the text. Performance contract: one append materializes at most one node and copies the projection only when it adds that node; an event that changes no node keeps the previous array reference (a chunk storm costs nothing), and unchanged nodes keep their object identity.
 
+## Request inspection
+
+`SessionHistoryInspection.requests` is one chronological, purpose-discriminated provider-request stream. Assistant requests always carry their numeric `turn` and `step`; compaction requests carry `step: 0` and a `turn` owner that may be `null`. That null owner means a manual compaction ran standalone between turns, not that it belongs to either adjacent turn. A `session/end-seed` boundary closes an unmatched compaction request as an error at the boundary time with `Compaction was interrupted before completion.`; a later start projects as an independent request instead of overwriting the orphan.
+
 ## Code Mode sub-dispatch index
 
 `ConversationSnapshot.codeDispatches` groups a `run_code` call's sub-dispatches under their parent callId, in start order, using the native call-block shapes: a `tool/code-dispatch-start` event lands the `RunningToolCall` form (rows derive the running ring from the shape) and its `tool/code-dispatch` settlement replaces it in place with the `ToolResultNode` form, `callTime` carrying the paired start's time. A settle whose start fell outside the replay window appends directly with `callTime: null` (duration unknown — never a fabricated zero). Live mux frames and history replay build the identical index; sub-calls never join the transcript `nodes` flow; per-parent array and map references are memo-stable across unrelated snapshot swaps.
@@ -40,7 +44,7 @@ Because the projection is log-ordered, the node array is seq-monotonic by constr
 
 ## Model retry projection
 
-The Session object validates plugin-owned, provider-routed `llm/retry` payloads at the event wire boundary against the producer's complete field contract, including timer, integer, status, provider-delay, and non-empty diagnostic bounds. A valid event removes the matching failed step's streaming partial and inserts a durable retry notice at the event's sequence position. The notice is `scheduled` until a following retry turn starts; an aborted or disposed source turn marks it `cancelled`, while the retry turn marks it `started`. Normal-mode notices carry their finite maximum; always-mode notices remain explicitly unbounded. Window rebuild and history replay apply the same projection, so logged chunks from the discarded attempt never reappear as an interrupted reply after refresh. A terminal turn without `llm/retry` retains the existing behavior: visible unfinalized output is frozen as an interrupted assistant node.
+The Session object validates plugin-owned, provider-routed `llm/retry` payloads at the event wire boundary against the producer's complete field contract, including timer, integer, status, provider-delay, and non-empty diagnostic bounds. A valid event removes the matching failed step's streaming partial and inserts a durable retry notice at the event's sequence position. The notice is `scheduled` until a following retry turn starts; an aborted or disposed source turn marks it `cancelled`, while the retry turn marks it `started`. Normal-mode notices carry their finite maximum; always-mode notices remain explicitly unbounded. A terminal `turn/end` error without a retry projects one `turn-error` node from its durable message and optional code; AUTH projections replace provider copy that may echo credential fragments with `API key is invalid`, while the raw diagnostic remains in the session log. A retried failure keeps only the retry notice for that attempt. Window rebuild and history replay apply the same projection, so refresh neither resurrects discarded chunks nor loses terminal failure feedback. Visible unfinalized output is frozen as an interrupted assistant node beside the terminal error.
 
 ## Session forking
 

packages/client/runtime/README.zh.md

@@ -30,6 +30,10 @@ SlotsService 分别为 renderer 提供 `useSessions` 与 `useWorkspaces` 的裸
 
 由于投影按日志顺序，节点数组天然按 seq 单调：仅日志的 `command/run` / `command/done` 节点按 seq 插入，`Session` 按分数 seq 归并被打断的冻结节点，而检查点所引范围落在窗口之外的窗口会渲染出标记且不打印任何日志。标记的摘要文本来自检查点的 `compact/summary` 溯源；窗口切分把溯源留在窗口外时该行不可展开而非空白，后续补上溯源的分页会解析出文本。性能契约：一次追加最多物化一个节点，并且仅在加入该节点时复制投影；不改变任何节点的事件保持上一次的数组引用（分片风暴零成本），未变化的节点保持其对象标识。
 
+## 请求检查
+
+`SessionHistoryInspection.requests` 是一条按时间顺序排列、以用途为判别字段的提供方请求流。助手请求始终携带数值型 `turn` 与 `step`；压缩请求携带 `step: 0`，其 `turn` 所有者可以是 `null`。这个 null 所有者表示手动压缩独立运行在两个轮次之间，并不表示它属于任一相邻轮次。`session/end-seed` 边界会在边界时刻将未匹配的压缩请求以错误状态结束，错误固定为 `Compaction was interrupted before completion.`；后续 start 会投影为独立请求，而不会覆盖这项遗留的未匹配请求。
+
 ## Code Mode 子调用索引
 
 `ConversationSnapshot.codeDispatches` 按父调用的 callId 和启动顺序，用原生调用块形状组织一个 `run_code` 调用的子调用：`tool/code-dispatch-start` 事件落成 `RunningToolCall` 形状（行组件从该形状推导运行中的转圈状态），其 `tool/code-dispatch` 完结事件原位替换为 `ToolResultNode` 形状，`callTime` 携带成对 start 事件的时间。start 落在回放窗口之外的完结事件则直接追加，`callTime: null`（耗时未知——绝不伪造零耗时）。live mux 帧与历史回放构建相同的索引；子调用永不进入 transcript 的 `nodes` 流；无关快照交换不会改变每个父调用对应的数组引用和映射引用，两者均保持 memo 稳定。
@@ -40,7 +44,7 @@ SlotsService 分别为 renderer 提供 `useSessions` 与 `useWorkspaces` 的裸
 
 ## 模型重试投影
 
-Session 对象会在事件 wire 边界依据生产方的完整字段契约，验证由插件负责、按提供方路由的 `llm/retry` 载荷，包括计时器、整数、状态、提供方延迟和非空诊断字段的边界。有效事件会移除对应失败步骤的流式输出片段，并在该事件的序列位置插入一条持久的重试提示。该提示在后续重试轮次开始前为 `scheduled`；源轮次中止或被 dispose（资源释放）时，会将该提示标记为 `cancelled`，重试轮次则会将其标记为 `started`。normal mode 提示携带其有限上限；always mode 提示则保持显式无界。窗口重建与历史回放应用相同的投影，因此刷新后，来自已丢弃尝试的日志分片绝不会重新显示为中断回复。没有 `llm/retry` 的终止轮次保留现有行为：可见但尚未定稿的输出会冻结为中断的 assistant 节点。
+Session 对象会在事件 wire 边界依据生产方的完整字段契约，验证由插件负责、按提供方路由的 `llm/retry` 载荷，包括计时器、整数、状态、提供方延迟和非空诊断字段的边界。有效事件会移除对应失败步骤的流式输出片段，并在该事件的序列位置插入一条持久的重试提示。该提示在后续重试轮次开始前为 `scheduled`；源轮次中止或被 dispose（资源释放）时，会将该提示标记为 `cancelled`，重试轮次则会将其标记为 `started`。normal mode 提示携带其有限上限；always mode 提示则保持显式无界。没有重试的终态 `turn/end` 错误会从持久消息与可选错误码投影出一个 `turn-error` 节点；AUTH 投影会把可能回显凭据片段的提供方文案替换为 `API key is invalid`，原始诊断仍保留在会话日志中。进入重试的失败则只保留该次尝试的重试提示。窗口重建与历史回放应用相同的投影，因此刷新既不会让已丢弃的分片重新出现，也不会丢失终态失败反馈。可见但尚未定稿的输出会在终态错误旁冻结为中断的 assistant 节点。
 
 ## 会话 fork
 

packages/client/runtime/src/client/index.ts

@@ -47,7 +47,7 @@ export type {
   AssistantTiming, CodeSubCall, CommandNode, CompactionSummaryNode, ComposerPhase,
   ContextMessageNode, ConversationNode, ConversationSnapshot, ModelRetryNode, QueuedMessage,
   RunningToolCall,
-  SteeringMessageNode, TodoItem, ToolResultNode, UnknownSurfaceNode, UserMessageNode,
+  SteeringMessageNode, TodoItem, ToolResultNode, TurnErrorNode, UnknownSurfaceNode, UserMessageNode,
 } from './sessions/conversation.ts'
 export type {
   ConversationContext, ConversationContextOriginKind,

packages/client/runtime/src/client/sessions/conversation.ts

@@ -138,6 +138,19 @@ export type ModelRetryNode = LlmRetryEventData & {
   retryState: 'scheduled' | 'started' | 'cancelled'
 }
 
+/** Durable terminal failure for a turn that has no scheduled retry. */
+export interface TurnErrorNode {
+  kind: 'turn-error'
+  /** Seq of the owning turn/end event. */
+  seq: number
+  /** Unix epoch ms from the turn/end event. */
+  time: number
+  turn: number
+  step: number
+  message: string
+  code?: string
+}
+
 /** A tool result paired (when in-window) with its call head. */
 export interface ToolResultNode {
   kind: 'tool-result'
@@ -226,6 +239,7 @@ export type ConversationNode =
   | SteeringMessageNode
   | ContextMessageNode
   | ModelRetryNode
+  | TurnErrorNode
   | ToolResultNode
   | CommandNode
   | CompactionSummaryNode

packages/client/runtime/src/client/sessions/failure-display.ts

@@ -0,0 +1,10 @@
+/**
+ * Convert a durable failure into copy that is safe to expose in the GUI.
+ * @param failure - Structured failure preserved by the session event.
+ * @returns Display-safe copy for client projections.
+ */
+export function displayFailureMessage(failure: { code?: string; message: string }): string {
+  // Provider AUTH messages may echo a masked or partially preserved credential.
+  // Keep the raw diagnostic in the session log, but never project it into UI state.
+  return failure.code === 'AUTH' ? 'API key is invalid' : failure.message
+}

packages/client/runtime/src/client/sessions/request-inspection.ts

@@ -8,6 +8,7 @@ import type { SessionEvent } from '@deepseek-ai/dsh-session/types'
 import type {
   AssistantProvenanceView, AssistantRequestConfig,
 } from './conversation.ts'
+import { displayFailureMessage } from './failure-display.ts'
 
 export type {
   AssistantProvenanceView, AssistantRequestConfig,
@@ -35,40 +36,55 @@ export interface RequestPromptChange {
   previous?: ConversationPromptSnapshot
 }
 
-/** One provider request reconstructed from durable request lifecycle events. */
-export interface RequestView {
-  /** Request category; compaction is a purpose, not a separate projection. */
-  purpose: 'assistant' | 'compaction'
+/** Lifecycle fields shared by ordinary generation and compaction requests. */
+interface RequestViewBase {
   /** Sequence that opened the operation represented by this request. */
   startSeq: number
-  turn: number
-  /** Agent-loop step, or zero for a direct compaction request. */
-  step: number
   startedAt: number
   completedAt: number | null
   status: 'running' | 'complete' | 'error'
   error?: string
-  /** Effective ordinary request input, inherited until a later header changes it. */
-  prompt?: ConversationPromptSnapshot
-  /** Prompt change logged while preparing this request. */
-  promptChange?: RequestPromptChange
   provenance?: AssistantProvenanceView
   requestConfig?: AssistantRequestConfig
   usage?: unknown
   /** Assistant message or compaction summary sequence produced by this request. */
   resultSeq?: number
+}
+
+/** One ordinary assistant generation reconstructed from durable request events. */
+interface AssistantRequestView extends RequestViewBase {
+  purpose: 'assistant'
+  turn: number
+  /** Agent-loop step that issued this request. */
+  step: number
+  /** Effective ordinary request input, inherited until a later header changes it. */
+  prompt?: ConversationPromptSnapshot
+  /** Prompt change logged while preparing this request. */
+  promptChange?: RequestPromptChange
+  /** Retry ordinal scheduled after a failed ordinary request. */
+  retry?: number
+  maxRetries?: number
+  retryDelayMs?: number
+}
+
+/** One compaction provider request, either turn-owned or standalone between turns. */
+interface CompactionRequestView extends RequestViewBase {
+  purpose: 'compaction'
+  /** Owning turn, or `null` when manual compaction ran between turns. */
+  turn: number | null
+  /** Direct compaction requests do not consume an agent-loop step. */
+  step: 0
   /** Compaction replacement message sequence, when one was committed. */
   replacementSeq?: number
   /** Safe compaction summary projection. */
   summary?: readonly ContentBlock[]
   /** Complete compaction provider output before the safe projection. */
   rawOutput?: readonly ContentBlock[]
-  /** Retry ordinal scheduled after a failed ordinary request. */
-  retry?: number
-  maxRetries?: number
-  retryDelayMs?: number
 }
 
+/** One provider request reconstructed from durable request lifecycle events. */
+export type RequestView = AssistantRequestView | CompactionRequestView
+
 /** Immutable request-centric projection derived from one history window. */
 export interface RequestInspectionSnapshot {
   requests: readonly RequestView[]
@@ -110,7 +126,7 @@ interface CompactionStartEvent {
   type: 'compact/start'
   seq: number
   time: number
-  data: { turn: number }
+  data: { turn: number | null }
 }
 
 interface CompactionSummaryEvent {
@@ -131,7 +147,7 @@ interface CompactionEndEvent {
   type: 'compact/end'
   seq: number
   time: number
-  data: { turn: number; error?: string }
+  data: { turn: number | null; error?: string }
 }
 
 function requestKey(turn: number, step: number): string {
@@ -228,10 +244,21 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
   let activePrompt: ConversationPromptSnapshot | undefined
   let activeCompaction: number | undefined
 
-  const update = (index: number | undefined, change: Partial<RequestView>): void => {
+  const updateAssistant = (
+    index: number | undefined,
+    change: Partial<Omit<AssistantRequestView, 'purpose'>>,
+  ): void => {
     if (index === undefined) return
     const request = requests[index]
-    if (request !== undefined) requests[index] = { ...request, ...change }
+    if (request?.purpose === 'assistant') requests[index] = { ...request, ...change }
+  }
+  const updateCompaction = (
+    index: number | undefined,
+    change: Partial<Omit<CompactionRequestView, 'purpose'>>,
+  ): void => {
+    if (index === undefined) return
+    const request = requests[index]
+    if (request?.purpose === 'compaction') requests[index] = { ...request, ...change }
   }
 
   for (const sourceEvent of events) {
@@ -263,7 +290,7 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
       }
       const change = promptChange(activePrompt, prompt, sourceEvent)
       activePrompt = prompt
-      update(activeStep === undefined ? undefined : ordinaryByStep.get(activeStep), {
+      updateAssistant(activeStep === undefined ? undefined : ordinaryByStep.get(activeStep), {
         prompt,
         requestConfig: prompt.config,
         ...(change === undefined ? {} : { promptChange: change }),
@@ -278,8 +305,11 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
         requestKey(sourceEvent.data.turn, sourceEvent.data.step),
       )
       const request = index === undefined ? undefined : requests[index]
-      update(index, {
-        usage: addTokenUsage(request?.usage, sourceEvent.data.chunk.usage),
+      updateAssistant(index, {
+        usage: addTokenUsage(
+          request?.purpose === 'assistant' ? request.usage : undefined,
+          sourceEvent.data.chunk.usage,
+        ),
       })
       continue
     }
@@ -288,7 +318,7 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
         requestKey(sourceEvent.data.turn, sourceEvent.data.step),
       )
       const request = index === undefined ? undefined : requests[index]
-      update(index, {
+      updateAssistant(index, {
         completedAt: sourceEvent.time,
         status: 'complete',
         resultSeq: sourceEvent.seq,
@@ -296,7 +326,9 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
           provider: sourceEvent.data.message.source.provider,
           model: sourceEvent.data.message.source.model,
         },
-        ...(request?.usage !== undefined || sourceEvent.data.usage === undefined
+        ...(request?.purpose === 'assistant'
+          && request.usage !== undefined
+          || sourceEvent.data.usage === undefined
           ? {}
           : { usage: sourceEvent.data.usage }),
       })
@@ -306,8 +338,8 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
       const key = requestKey(sourceEvent.data.turn, sourceEvent.data.step)
       const index = ordinaryByStep.get(key)
       const request = index === undefined ? undefined : requests[index]
-      if (request?.status === 'running') {
-        update(index, {
+      if (request?.purpose === 'assistant' && request.status === 'running') {
+        updateAssistant(index, {
           completedAt: sourceEvent.time,
           status: 'error',
         })
@@ -317,9 +349,9 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
     }
     if ((sourceEvent.type as string) === 'llm/retry') {
       const event = sourceEvent as unknown as RetryEvent
-      update(ordinaryByStep.get(requestKey(event.data.turn, event.data.step)), {
+      updateAssistant(ordinaryByStep.get(requestKey(event.data.turn, event.data.step)), {
         status: 'error',
-        error: event.data.failure.message,
+        error: displayFailureMessage(event.data.failure),
         retry: event.data.retry,
         maxRetries: event.data.maxRetries,
         retryDelayMs: event.data.delayMs,
@@ -328,14 +360,23 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
     }
     if (sourceEvent.type === 'turn/end' && sourceEvent.data.reason.kind === 'error') {
       const reason = sourceEvent.data.reason
-      update(ordinaryByStep.get(requestKey(sourceEvent.data.turn, reason.step)), {
+      updateAssistant(ordinaryByStep.get(requestKey(sourceEvent.data.turn, reason.step)), {
         status: 'error',
-        error: 'failure' in reason ? reason.failure.message : reason.message,
+        error: displayFailureMessage('failure' in reason ? reason.failure : reason),
       })
       continue
     }
 
     const type = sourceEvent.type as string
+    if (type === 'session/end-seed' && activeCompaction !== undefined) {
+      updateCompaction(activeCompaction, {
+        completedAt: sourceEvent.time,
+        status: 'error',
+        error: 'Compaction was interrupted before completion.',
+      })
+      activeCompaction = undefined
+      continue
+    }
     if (type === 'compact/start') {
       const event = sourceEvent as unknown as CompactionStartEvent
       activeCompaction = requests.length
@@ -352,7 +393,7 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
     }
     if (type === 'compact/summary' && activeCompaction !== undefined) {
       const event = sourceEvent as unknown as CompactionSummaryEvent
-      update(activeCompaction, {
+      updateCompaction(activeCompaction, {
         resultSeq: event.seq,
         summary: event.data.summary,
         ...(event.data.rawOutput === undefined ? {} : { rawOutput: event.data.rawOutput }),
@@ -375,12 +416,12 @@ function deriveRequests(events: readonly SessionEvent[]): readonly RequestView[]
       && activeCompaction !== undefined
       && isCompactionSource(sourceEvent.data.source)
     ) {
-      update(activeCompaction, { replacementSeq: sourceEvent.seq })
+      updateCompaction(activeCompaction, { replacementSeq: sourceEvent.seq })
       continue
     }
     if (type !== 'compact/end' || activeCompaction === undefined) continue
     const event = sourceEvent as unknown as CompactionEndEvent
-    update(activeCompaction, {
+    updateCompaction(activeCompaction, {
       completedAt: event.time,
       status: event.data.error === undefined ? 'complete' : 'error',
       ...(event.data.error === undefined ? {} : { error: event.data.error }),

packages/client/runtime/src/client/sessions/session.ts

@@ -20,6 +20,7 @@ import type {
 import type { PendingInteraction } from './pending.ts'
 import { PendingWait } from './pending.ts'
 import { TranscriptAdapter } from './transcript-adapter.ts'
+import { displayFailureMessage } from './failure-display.ts'
 import { Notifier } from './notifier.ts'
 import { PartialAccumulator } from './partial.ts'
 import { ProjectionValueStore } from './projection-store.ts'
@@ -772,6 +773,22 @@ export class Session implements SessionFace {
         if (event.data.reason.kind === 'aborted' || event.data.reason.kind === 'disposed') {
           this.settleScheduledRetry('cancelled', event.data.turn)
         }
+        if (
+          event.data.reason.kind === 'error'
+          && !this.derivedNodes.some(node => node.kind === 'model-retry' && node.turn === event.data.turn)
+        ) {
+          const failure = 'failure' in event.data.reason ? event.data.reason.failure : event.data.reason
+          this.derivedNodes.push({
+            kind: 'turn-error',
+            seq: event.seq,
+            time: event.time,
+            turn: event.data.turn,
+            step: event.data.reason.step,
+            message: displayFailureMessage(failure),
+            ...(failure.code === undefined ? {} : { code: failure.code }),
+          })
+          this.derivedRev++
+        }
         // Aborted turns never finalize. The accumulated partial is VALUE, not residue: freeze it
         // into an interrupted terminal node (pulse stops, text survives) instead of deleting it.
         // Shared by live and window-replay paths, so a refresh reconstructs the same frozen node

packages/client/runtime/tests/request-inspection.spec.ts

@@ -85,6 +85,68 @@ describe('inspectRequests', () => {
     expect(snapshot.callSchemas.get('call-1')?.name).toBe('read')
   })
 
+  it('preserves a standalone compaction owner without widening assistant turns', () => {
+    const snapshot = inspectRequests(entriesOf([
+      at(0, 'compact/start', { turn: null }),
+      at(1, 'compact/summary', {
+        summary: [{ type: 'text', text: 'standalone summary' }],
+        provider: 'fake',
+        model: 'compact-model',
+      }),
+      at(2, 'compact/end', { turn: null }),
+      at(3, 'step/start', { turn: 2, step: 1 }),
+    ]))
+
+    const [compaction, assistant] = snapshot.requests
+    expect(compaction).toMatchObject({
+      purpose: 'compaction',
+      turn: null,
+      step: 0,
+      status: 'complete',
+    })
+    expect(assistant).toMatchObject({
+      purpose: 'assistant',
+      turn: 2,
+      step: 1,
+      status: 'running',
+    })
+    if (assistant?.purpose === 'assistant') {
+      const turn: number = assistant.turn
+      expect(turn).toBe(2)
+    }
+  })
+
+  it('interrupts an orphaned compaction at end-seed before projecting a new attempt', () => {
+    const snapshot = inspectRequests(entriesOf([
+      at(0, 'compact/start', { turn: null }),
+      at(1, 'session/end-seed', {}),
+      at(2, 'compact/start', { turn: null }),
+      at(3, 'compact/summary', {
+        summary: [{ type: 'text', text: 'replacement summary' }],
+        provider: 'fake',
+        model: 'compact-model',
+      }),
+      at(4, 'compact/end', { turn: null }),
+    ]))
+
+    expect(snapshot.requests).toMatchObject([
+      {
+        purpose: 'compaction',
+        startSeq: 0,
+        status: 'error',
+        completedAt: 1_700_000_000_001,
+        error: 'Compaction was interrupted before completion.',
+      },
+      {
+        purpose: 'compaction',
+        startSeq: 2,
+        status: 'complete',
+        completedAt: 1_700_000_000_004,
+        summary: [{ type: 'text', text: 'replacement summary' }],
+      },
+    ])
+  })
+
   it('captures schemas for nested tool dispatches from the active request header', () => {
     const snapshot = inspectRequests(entriesOf([
       at(0, 'request/header', {
@@ -159,6 +221,33 @@ describe('inspectRequests', () => {
     })
   })
 
+  it('keeps provider credential fragments out of projected request errors', () => {
+    const snapshot = inspectRequests(entriesOf([
+      at(0, 'step/start', { turn: 1, step: 1 }),
+      at(1, 'turn/end', {
+        turn: 1,
+        reason: {
+          kind: 'error',
+          step: 1,
+          failure: {
+            code: 'AUTH',
+            message: 'Authentication Fails, Your api key: sk-preview-secret is invalid',
+          },
+        },
+      }),
+      at(2, 'step/start', { turn: 2, step: 1 }),
+      at(3, 'turn/end', {
+        turn: 2,
+        reason: { kind: 'error', step: 1, message: 'plugin exploded' },
+      }),
+    ]))
+
+    expect(snapshot.requests).toMatchObject([
+      { status: 'error', error: 'API key is invalid' },
+      { status: 'error', error: 'plugin exploded' },
+    ])
+  })
+
   it('treats a scrubbed durable-fixture tool catalog as unavailable', () => {
     const snapshot = inspectRequests(entriesOf([
       at(0, 'step/start', { turn: 1, step: 1 }),
@@ -179,6 +268,7 @@ describe('inspectRequests', () => {
     ]))
 
     expect(snapshot.callSchemas).toEqual(new Map())
-    expect(snapshot.requests[0]?.prompt?.tools).toEqual([])
+    const [request] = snapshot.requests
+    expect(request?.purpose === 'assistant' ? request.prompt?.tools : undefined).toEqual([])
   })
 })

packages/client/runtime/tests/session.spec.ts

@@ -211,6 +211,7 @@ describe('live event path', () => {
     for (const event of retryTurn.slice(7)) feed(event)
     snapshot = session.getSnapshot()
     expect(snapshot.nodes.slice(-2).map(node => node.kind)).toEqual(['model-retry', 'assistant'])
+    expect(snapshot.nodes.some(node => node.kind === 'turn-error')).toBe(false)
     expect(snapshot.nodes.at(-2)).toMatchObject({ kind: 'model-retry', retryState: 'started' })
     expect(snapshot.nodes.at(-1)).toMatchObject({ kind: 'assistant', blocks: [{ kind: 'text', text: '完整回复' }] })
 
@@ -221,6 +222,50 @@ describe('live event path', () => {
     expect(replay.session.getSnapshot().partial).toBeNull()
   })
 
+  it('projects unretried terminal failures at turn/end and reproduces them from history', async () => {
+    const { session } = await opened()
+    const feed = (event: SessionEvent) => {
+      session.handleMuxEnvelope('r' as never, { type: 'session/event', sessionId: SID, event })
+    }
+    const failedTurns = [
+      ev.turnStart(6, 1),
+      ev.user(7, '鉴权失败'),
+      at(8, {
+        type: 'turn/end',
+        data: {
+          turn: 1,
+          reason: {
+            kind: 'error',
+            step: 0,
+            failure: {
+              code: 'AUTH',
+              message: 'Authentication Fails, Your api key: sk-preview-secret is invalid',
+            },
+          },
+        },
+      }),
+      ev.turnStart(9, 2),
+      ev.user(10, '内部失败'),
+      at(11, {
+        type: 'turn/end',
+        data: { turn: 2, reason: { kind: 'error', step: 1, message: 'plugin exploded' } },
+      }),
+    ]
+    for (const event of failedTurns) feed(event)
+
+    const errors = session.getSnapshot().nodes.filter(node => node.kind === 'turn-error')
+    expect(errors).toMatchObject([
+      { seq: 8, turn: 1, step: 0, code: 'AUTH', message: 'API key is invalid' },
+      { seq: 11, turn: 2, step: 1, message: 'plugin exploded' },
+    ])
+    expect('code' in errors[1]!).toBe(false)
+
+    const replay = makeSession()
+    replay.api.onHistory = () => histResponse([...plainTurn(0, 0, 'a', 'b'), ...failedTurns])
+    await replay.session.open()
+    expect(replay.session.getSnapshot().nodes).toEqual(session.getSnapshot().nodes)
+  })
+
   it('rejects retry payloads outside the producer contract without retracting the current partial', async () => {
     const { session } = await opened()
     const feed = (event: SessionEvent) => { session.handleMuxEnvelope('r' as never, { type: 'session/event', sessionId: SID, event }) }

packages/client/ui-conversation/README.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 packages/client/ui-conversation/README.md
-README.md: 10c1fdcdfbe9d7bfdf5f19767d7898ffc5163f99
-README.zh.md: db3a20bfd0b318e1f721e6d374f39206e7565adb
+README.md: 2dafe3c69bb087550f6c0f7faab12bff1f1324ac
+README.zh.md: 4201441c6c6397b7b35fe0fab99fc6fc6271719f

packages/client/ui-conversation/README.md

@@ -24,7 +24,7 @@ A `read` call declaring the `read` render intent renders the returned file windo
 
 A tool call declaring the `diff` render intent (the `write`/`edit` tools) renders its applied change inline through ui-primitives' `DiffBlock`, the same four-layer shape. `contract/diff-card-model.ts` is the single derivation from the `callView`/`resultView` pair; the settled result's hunks replace the call-time diff, and it yields null — the generic path — for any other card tag or a generic result view (write/edit's execution errors). The keyed `FileMutationRow` (registered under both `write` and `edit`) composes the shared `ToolRow`, feeding the diff as ToolRow's `diff` body, so it is the row's collapsed-by-default expanded card; the summary path link still opens the file through the host, and an errored mutation (no diff card) surfaces its error text through ToolRow's Output section with the first line in the collapsed summary. The render-site fallback and the details panel are diff-aware too. Rows cap at `CHAT_DIFF_MAX_LINES` (8) against the panel's 16 ([decision](../../../.agents/notes/implemented/feature/2026-07-30-web-diff-card.md)).
 
-The chat flow projects consecutive model-retry nodes across retry turns into one stable, muted status row updated to the latest attempt; every retry event remains in the runtime snapshot and session log. Its frontend countdown anchors the scheduled delay to client receipt, avoiding host/browser clock skew, rounds remaining time up to seconds, and has a one-second floor. The latest unresolved retry uses a left-to-right text shimmer. Subsequent turn facts distinguish an attempt that started from one cancelled during backoff, while the Host running bit only controls the live animation; the row then shows a static completed or cancelled label. Normal policy rows show the finite retry maximum; always policy rows show `∞`. Activating the row reveals the latest exact retry delay and failure message. The client runtime removes each failed step's streaming tail before its retry node arrives, while the status remains visible after a later attempt succeeds.
+The chat flow projects consecutive model-retry nodes across retry turns into one stable, muted status row updated to the latest attempt; every retry event remains in the runtime snapshot and session log. Its frontend countdown anchors the scheduled delay to client receipt, avoiding host/browser clock skew, rounds remaining time up to seconds, and has a one-second floor. The latest unresolved retry uses a left-to-right text shimmer. Subsequent turn facts distinguish an attempt that started from one cancelled during backoff, while the Host running bit only controls the live animation; the row then shows a static completed or cancelled label. Normal policy rows show the finite retry maximum; always policy rows show `∞`. Activating the row reveals the latest exact retry delay and failure message. The client runtime removes each failed step's streaming tail before its retry node arrives, while the status remains visible after a later attempt succeeds. An unretried terminal failure renders as a persistent inline status at its turn boundary, showing the display-safe durable message and optional error code without offering an action the Host cannot fulfill; AUTH copy never echoes provider-supplied credential fragments.
 
 A `grep`/`glob` call declaring the `search` render intent renders its result inline, at the same render sites, through ui-primitives' `SearchBlock` — grep's matches grouped by file (each a collapsible header of `lineNumber: line` rows), glob's flat path list. `contract/search-card-model.ts` is the single derivation from the snapshot's `resultView`; unlike the terminal card it reads no `callView`, since a search has no matches or paths before `execute`, so a running search shows its summary alone. It yields null — the generic path — for any non-search result view, a `card` or `kind` this client version does not compile, and (because those ride the untrusted wire frame) a known kind whose `files`/`paths` is malformed. The keyed `SearchRow`, registered under both `grep` and `glob` since the derived `kind` decides the shape, composes the shared `ToolRow`, feeding the card as ToolRow's `search` body, so it is the row's collapsed-by-default expanded card; the render-site fallback routes it the same way. Both cap at `CHAT_SEARCH_MAX_LINES` (8) against the panel's 16. A capped search drops rows from the card, but the locator to the rest — grep/glob's `Full … stored at …` footer — lives only in the result text, so the derivation surfaces that as a recovery footer below the card when (and only when) the result was truncated; a settled call with no card at all (an errored search, a nested `run_code` sub-dispatch, a legacy generic result) surfaces its flattened result text through ToolRow's Output section so nothing is lost behind a bare summary ([decision](../../../.agents/notes/implemented/feature/2026-07-30-web-search-card.md)).
 

packages/client/ui-conversation/README.zh.md

@@ -22,7 +22,7 @@
 
 声明 `diff` 渲染意图的工具调用（`write`／`edit` 工具），通过 ui-primitives 的 `DiffBlock` 内联渲染其已应用的改动，采用同一套四层结构。`contract/diff-card-model.ts` 是从 `callView`／`resultView` 对推导的唯一位置；已结算 result 的 hunk 替换 call 时 diff，对任何其他 card 标签或 generic result view（write/edit 的执行错误）它返回 null，落回通用路径。键控的 `FileMutationRow`（在 `write` 与 `edit` 下都注册）组合共享的 `ToolRow`，把 diff 作为 ToolRow 的 `diff` body 传入，因此它是该行默认折叠的展开卡片；摘要路径链接仍经 host 打开文件，而出错的改动（没有 diff 卡片）经 ToolRow 的 Output 区呈现其错误文本，首行进入折叠摘要。渲染点兜底行与详情面板同样感知 diff。行的上限是 `CHAT_DIFF_MAX_LINES`（8），面板为 16（[决策](../../../.agents/notes/implemented/feature/2026-07-30-web-diff-card.md)）。
 
-聊天流会将跨重试轮次连续出现的模型重试节点投影为一个稳定的弱化状态行，并用最新一次尝试更新该行；每个重试事件仍保留在运行时快照与会话日志中。前端倒计时以客户端收到事件的时刻为计划延迟的起点，避免 Host 与浏览器的时钟偏差；剩余时间向上取整到秒，且下限为 1 秒。最近一次尚未完成的重试会显示从左到右的文字渐变动画。后续轮次事实用于区分已开始的尝试与在退避期间取消的尝试，Host 的 running 位只控制实时动画；随后该行会显示静态的已完成或已取消标签。normal 策略行显示有限重试上限；always 策略行显示 `∞`。激活该行会显示最近一次重试的精确延迟和失败消息。客户端运行时会在相应重试节点到达前移除每个失败步骤的流式输出尾部；后续某次尝试成功后，该状态仍保持可见。
+聊天流会将跨重试轮次连续出现的模型重试节点投影为一个稳定的弱化状态行，并用最新一次尝试更新该行；每个重试事件仍保留在运行时快照与会话日志中。前端倒计时以客户端收到事件的时刻为计划延迟的起点，避免 Host 与浏览器的时钟偏差；剩余时间向上取整到秒，且下限为 1 秒。最近一次尚未完成的重试会显示从左到右的文字渐变动画。后续轮次事实用于区分已开始的尝试与在退避期间取消的尝试，Host 的 running 位只控制实时动画；随后该行会显示静态的已完成或已取消标签。normal 策略行显示有限重试上限；always 策略行显示 `∞`。激活该行会显示最近一次重试的精确延迟和失败消息。客户端运行时会在相应重试节点到达前移除每个失败步骤的流式输出尾部；后续某次尝试成功后，该状态仍保持可见。未进入重试的终态失败会在其轮次边界渲染为持久的内联状态，展示适合显示的持久消息与可选错误码，但不会提供 Host 无法兑现的操作；AUTH 文案绝不会回显提供方给出的凭据片段。
 
 声明 `search` 渲染意图的 `grep`／`glob` 调用，会在同样的渲染点上通过 ui-primitives 的 `SearchBlock` 内联渲染其结果——grep 的匹配按文件分组（每个是一个可折叠的头，下辖 `lineNumber: line` 行），glob 是扁平路径列表。`contract/search-card-model.ts` 是从快照的 `resultView` 推导的唯一位置；与终端卡片不同，它不读 `callView`，因为搜索在 `execute` 前没有匹配或路径，所以运行中的搜索只显示摘要。对任何非搜索的结果视图、当前客户端版本无法编译的 `card` 或 `kind`、以及（因为这些都与不可信的 wire 帧同行）一个 `files`／`paths` 格式错误的已知 kind，它都返回 null，落回通用路径。键控的 `SearchRow` 因推导出的 `kind` 决定形态而同时注册在 `grep` 与 `glob` 下，组合共享的 `ToolRow`，把卡片作为 ToolRow 的 `search` body 传入，因此它是该行默认折叠的展开卡片；渲染点兜底行以同样方式渲染它。两者上限都是 `CHAT_SEARCH_MAX_LINES`（8），面板为 16。被截断的搜索会从卡片里丢掉一些行，但通往其余部分的定位符——grep/glob 的 `Full … stored at …` 脚注——只存在于结果文本里，因此推导在（且仅在）结果被截断时把它作为恢复脚注画在卡片下方；一个完全没有卡片的已结算调用（出错的搜索、嵌套 `run_code` 子派发、旧日志的 generic 结果）则经 ToolRow 的 Output 区呈现其压平后的结果文本，从而不让任何内容丢失在一个光秃秃的摘要之后（[决策](../../../.agents/notes/implemented/feature/2026-07-30-web-search-card.md)）。
 

packages/client/ui-conversation/src/client/chat/MessageItem.module.css

@@ -200,6 +200,40 @@
   color: var(--dsw-alias-label-secondary);
 }
 
+.turnErrorRow {
+  display: grid;
+  grid-template-columns: 10px minmax(0, 1fr) auto;
+  gap: 8px;
+  align-items: start;
+  padding: 2px 0;
+  font-size: 13px;
+  line-height: 20px;
+}
+
+.turnErrorDot {
+  margin-top: 5px;
+}
+
+.turnErrorCopy {
+  min-width: 0;
+  overflow-wrap: anywhere;
+}
+
+.turnErrorTitle {
+  margin-right: 6px;
+  color: var(--dsw-alias-state-error-primary);
+  font-weight: 600;
+}
+
+.turnErrorMessage {
+  color: var(--dsw-alias-label-secondary);
+}
+
+.turnErrorCode {
+  color: var(--dsw-alias-label-tertiary);
+  font: var(--dsw-font-markdown-code-block-small);
+}
+
 @keyframes retry-shimmer {
   from {
     background-position: 100% 50%;

packages/client/ui-conversation/src/client/chat/MessageItem.tsx

@@ -6,9 +6,9 @@ import { memo, useEffect, useMemo, useState } from 'react'
 import type { ReactNode } from 'react'
 import type {
   CompactionSummaryNode, ContextMessageNode, ModelRetryNode, SteeringMessageNode,
-  UnknownSurfaceNode, UserMessageNode,
+  TurnErrorNode, UnknownSurfaceNode, UserMessageNode,
 } from '@deepseek-ai/dsh-client-runtime/client'
-import { JsonBlock, MessageText } from '@deepseek-ai/dsh-client-ui-primitives'
+import { JsonBlock, MessageText, StateDot } from '@deepseek-ai/dsh-client-ui-primitives'
 import type { ChatViewSlotProps } from '../contract/slots.ts'
 import { CompactionItem } from './CompactionItem.tsx'
 import { ContextInjectionRow } from './ContextInjectionRow.tsx'
@@ -17,7 +17,14 @@ import css from './MessageItem.module.css'
 import { ImageGallery, type ImageLoader } from './MessageImage.tsx'
 
 export interface MessageItemProps {
-  node: UserMessageNode | SteeringMessageNode | ContextMessageNode | CompactionSummaryNode | ModelRetryNode | UnknownSurfaceNode
+  node:
+    | UserMessageNode
+    | SteeringMessageNode
+    | ContextMessageNode
+    | CompactionSummaryNode
+    | ModelRetryNode
+    | TurnErrorNode
+    | UnknownSurfaceNode
   loadImage?: ImageLoader
   retryActive?: boolean
   /** Fork the session through the turn containing this message (user-bubble branch action). */
@@ -121,6 +128,24 @@ function ModelRetryItem({ node, active, t }: {
     </details>
   )
 }
+
+/** Persistent, turn-positioned feedback for a terminal failure. */
+function TurnErrorItem({ node, t }: {
+  node: TurnErrorNode
+  t: ChatViewSlotProps['t']
+}) {
+  return (
+    <div className={css.turnErrorRow} role="status">
+      <StateDot state="error" className={css.turnErrorDot} />
+      <div className={css.turnErrorCopy}>
+        <span className={css.turnErrorTitle}>{t('message.turnError')}</span>
+        <span className={css.turnErrorMessage}>{node.message}</span>
+      </div>
+      {node.code !== undefined && <code className={css.turnErrorCode}>{node.code}</code>}
+    </div>
+  )
+}
+
 /**
  * Display projection of reference forms in a user bubble (free geometry — no
  * textarea alignment constraint here); everything else stays plain text. The
@@ -214,6 +239,8 @@ export const MessageItem = memo(function MessageItem({
       return <CompactionItem node={node} t={t} />
     case 'model-retry':
       return <ModelRetryItem node={node} active={retryActive} t={t} />
+    case 'turn-error':
+      return <TurnErrorItem node={node} t={t} />
     default:
       return (
         <div className={css.contextRow}>

packages/client/ui-conversation/src/client/locales.ts

@@ -77,6 +77,7 @@ export const zh = {
   'message.retry.status': '{label}（{retry}/{maximum}） · {seconds}s',
   'message.retry.delay': '重试延迟：',
   'message.retry.failure': '失败原因：',
+  'message.turnError': '本轮运行失败',
   'command.running': '执行中…',
   'command.failed': '命令失败',
   'command.done': '已完成',
@@ -191,6 +192,7 @@ export const en = {
   'message.retry.status': '{label} ({retry}/{maximum}) · {seconds}s',
   'message.retry.delay': 'Retry delay: ',
   'message.retry.failure': 'Failure reason: ',
+  'message.turnError': 'This turn failed',
   'command.running': 'Running…',
   'command.failed': 'Command failed',
   'command.done': 'Completed',

packages/client/ui-conversation/src/client/skeleton/ConversationRoot.tsx

@@ -22,6 +22,7 @@ export function ConversationRoot({
   const session = useSession(s => s)
   const inputState = useInput(s => s)
   const cwd = useSessions(s => sessionId === undefined ? undefined : s.byId[sessionId]?.cwd)
+  const summaryBlank = useSessions(s => sessionId === undefined ? undefined : s.byId[sessionId]?.blank)
   const workspaces = useWorkspaces(s => s)
 
   const [pickerOpen, setPickerOpen] = useState(false)
@@ -65,8 +66,16 @@ export function ConversationRoot({
   // While a session is still replaying (loading + blank) the hero/docked
   // choice is unknowable — render the composer hidden instead of flashing
   // the centered hero and snapping to the docked bar (or vice versa).
+  // Exemption: a session the list summary already proves blank can only
+  // land on the hero, so hiding would blank the column for the whole
+  // history round-trip (the startup auto-selection flash) for nothing.
+  // The exemption is deliberately open-state-wide, not loading-only: a
+  // summary-blank session is the hero before its open starts (`cold`) and
+  // after one fails (`error`) for the same reason — there is no history.
   const settling = sessionId !== undefined && composerPhase === 'blank' && openState === 'loading'
-  const hero = sessionId === undefined || (composerPhase === 'blank' && openState === 'open')
+    && summaryBlank !== true
+  const hero = sessionId === undefined
+    || (composerPhase === 'blank' && (openState === 'open' || summaryBlank === true))
   const zone: InputZone | undefined =
     session === undefined || inputState === undefined ? undefined : { session, input: inputState }
 

packages/client/ui-conversation/tests/chat-view.spec.tsx

@@ -8,7 +8,7 @@ import { Profiler } from 'react'
 import { act, cleanup, fireEvent, render, within } from '@testing-library/react'
 import type {
   AssistantMessageNode, CommandNode, ConversationNode, ConversationSnapshot,
-  ModelRetryNode, RunningToolCall, SessionId, SessionListState, ToolResultNode,
+  ModelRetryNode, RunningToolCall, SessionId, SessionListState, ToolResultNode, TurnErrorNode,
   UserMessageNode, WorkspaceListState,
 } from '@deepseek-ai/dsh-client-runtime/client'
 import { bindSnapshotSelector } from '@deepseek-ai/dsh-client-web-react'
@@ -75,6 +75,11 @@ const retry = (seq: number): ModelRetryNode => ({
   retry: 1, maxRetries: 2, delayMs: 450,
   failure: { code: 'TRANSPORT', message: '连接被重置' },
 })
+const turnError = (seq: number, code?: string): TurnErrorNode => ({
+  kind: 'turn-error', seq, time: seq * 1_000, turn: 1, step: 0,
+  message: seq === 2 ? 'API key is invalid' : 'plugin exploded',
+  ...(code === undefined ? {} : { code }),
+})
 const toolResult = (seq: number, callId: string, name = 'bash'): ToolResultNode => ({
   kind: 'tool-result', seq, time: seq * 1_000, callId,
   call: { name, argsRaw: `{"command":"cmd-${callId}","description":"run ${callId}"}` },
@@ -295,6 +300,16 @@ describe('ChatView', () => {
     expect(within(cancelledDisclosure).getByRole('status').textContent).toContain('重试已取消')
   })
 
+  it('renders terminal turn failures inline with their durable message and optional code', () => {
+    const h = makeHarness({ nodes: [user(1, 'try'), turnError(2, 'AUTH'), turnError(3)] })
+    const view = render(<h.ChatView {...h.props} />)
+    const statuses = view.getAllByRole('status')
+    expect(statuses.map(status => status.textContent)).toEqual([
+      '本轮运行失败API key is invalidAUTH',
+      '本轮运行失败plugin exploded',
+    ])
+  })
+
   it('the expanded row Inspect pill hands the call id to inspectCall', () => {
     const h = makeHarness({
       nodes: [toolResult(3, 'a')],

packages/client/ui-conversation/tests/skeleton.spec.tsx

@@ -80,16 +80,25 @@ function mount(
   snapshot: ConversationSnapshot,
   workspaceRows: WorkspaceView[] = [{ ...workspace('one'), sessionIds: [SID] }],
   retargetWorkspace = vi.fn(async (_workspaceId: WorkspaceId) => {}),
-  /** When true, mimic overlay:true chain siblings (hidden fallback + takeover). */
-  overlayTakeover = false,
+  options: {
+    /** When true, mimic overlay:true chain siblings (hidden fallback + takeover). */
+    overlayTakeover?: boolean
+    /** The session list summary's `blank` flag — independent of the snapshot's. */
+    summaryBlank?: boolean
+    /** Drop the session's summary row entirely (a session the list has not caught up with). */
+    omitSummaryRow?: boolean
+  } = {},
 ) {
   const root = sid('root')
+  const rootRow = { id: root, displayTitle: 'Root', running: false, waitingApproval: false, blank: false, updatedAt: 1 }
+  const childRow = {
+    id: SID, displayTitle: 'Child', parentId: root, cwd: '/projects/one',
+    running: false, waitingApproval: false, blank: options.summaryBlank ?? false, updatedAt: 2,
+  }
+  const listed = options.omitSummaryRow !== true
   const sessions = createSnapshotStore<SessionListState>({
-    ids: [root, SID],
-    byId: {
-      [root]: { id: root, displayTitle: 'Root', running: false, waitingApproval: false, blank: false, updatedAt: 1 },
-      [SID]: { id: SID, displayTitle: 'Child', parentId: root, cwd: '/projects/one', running: false, waitingApproval: false, blank: false, updatedAt: 2 },
-    },
+    ids: listed ? [root, SID] : [root],
+    byId: { [root]: rootRow, ...listed && { [SID]: childRow } },
     current: SID,
     phase: 'ready',
   })
@@ -171,7 +180,7 @@ function mount(
     return <div data-testid={`view-${opts?.only ?? key}`} />
   }) as ConversationRootProps['renderSlot']
   const renderSlotChain = ((_key, _owner, opts) => (
-    overlayTakeover
+    options.overlayTakeover === true
       ? (
         <>
           <div data-chain-overlay-fallback="conversation.composer" style={{ display: 'none' }}>
@@ -233,7 +242,7 @@ describe('ConversationRoot resident composer', () => {
   })
 
   it('sticky composer seat wraps the whole overlay chain, not only the fallback stack', () => {
-    const b = mount(conversationSnapshot(), undefined, undefined, true)
+    const b = mount(conversationSnapshot(), undefined, undefined, { overlayTakeover: true })
     const seat = b.view.container.querySelector('[data-composer-seat]')
     const takeover = b.view.getByTestId('composer-takeover')
     const fallback = b.view.container.querySelector('[data-chain-overlay-fallback="conversation.composer"]')
@@ -274,6 +283,39 @@ describe('ConversationRoot resident composer', () => {
     expect(b.view.getByText('Selected Folder')).toBeTruthy()
   })
 
+  it('settling phase: a summary that does not prove the session blank hides the composer while it opens', () => {
+    const b = mount(conversationSnapshot({ composerPhase: 'blank', blank: true, openState: 'loading' }))
+    const root = b.view.container.querySelector('[data-phase]')
+    expect(root?.getAttribute('data-phase')).toBe('settling')
+    expect(b.view.queryByText('开始构建吧')).toBeNull()
+  })
+
+  it('settling phase: a session the list has no row for settles conservatively', () => {
+    const b = mount(
+      conversationSnapshot({ composerPhase: 'blank', blank: true, openState: 'loading' }),
+      undefined,
+      undefined,
+      { omitSummaryRow: true },
+    )
+    const root = b.view.container.querySelector('[data-phase]')
+    expect(root?.getAttribute('data-phase')).toBe('settling')
+  })
+
+  it('startup auto-selection: a summary-proven blank session opens straight into the hero', () => {
+    const b = mount(
+      conversationSnapshot({ composerPhase: 'blank', blank: true, openState: 'loading' }),
+      undefined,
+      undefined,
+      { summaryBlank: true },
+    )
+    // The summary already proves the outcome, so the settling hide would only
+    // blank the column for the history round-trip.
+    const root = b.view.container.querySelector('[data-phase]')
+    expect(root?.getAttribute('data-phase')).toBe('hero')
+    expect(b.view.getByText('开始构建吧')).toBeTruthy()
+    expect(b.view.getByRole('textbox')).toBeTruthy()
+  })
+
   it('same textarea DOM node survives the hero → active flip into the sticky scrollport', () => {
     const b = mount(conversationSnapshot({ composerPhase: 'blank', blank: true }))
     const before = b.view.getByRole('textbox')

packages/client/ui-trajectory/README.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 packages/client/ui-trajectory/README.md
-README.md: b9c8b849b3454fe46e1fc37713d9d3b9449734cf
-README.zh.md: 6ddc32f2f27c93f8ccc80b3d9b31d56d3cf4dd94
+README.md: a65c11aed9dd74f9b0b60795441f876c1d64b3ad
+README.zh.md: 6e25d24c6b65673b3d003e624b6e0727be60c0e1