fix(web): restrict message forks to completed turn tails

.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.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/2026-08-02-message-fork-actions-require-completed-turn-tail.md
+2026-08-02-message-fork-actions-require-completed-turn-tail.md: 1f5ee6b7ff37709e07544f3e7c73ba1eee951f8d
+2026-08-02-message-fork-actions-require-completed-turn-tail.zh.md: ed897e9af56a601c8fc42379e3907e3551c6389e

.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md

@@ -0,0 +1,27 @@
+# Agent Note: Message fork actions require a completed turn tail
+
+Status: implemented
+
+English | [中文](2026-08-02-message-fork-actions-require-completed-turn-tail.zh.md)
+
+## Problem
+
+The Web conversation attached branch to the last assistant node with nonempty text in each turn. A later tool result, interrupted reasoning node, or terminal error did not take ownership because those rows have no content-text IconActions. The branch icon could therefore appear beneath an assistant response while more rows from the same turn remained below it. The Host correctly expanded that message anchor through the containing `turn/end`, but the placement made the action look like a message-level cut and the child visibly inherited the same-turn suffix.
+
+## Decision
+
+`ConversationSnapshot.turnEnds` retains the completed turn boundaries present in the raw event window. The conversation view walks transcript nodes through each boundary and exposes branch only when the boundary's last node is a user message or a content-bearing assistant message. Open turns have no eligible message, and a later tool result, reasoning-only interruption, turn error, or other transcript node suppresses branch on earlier messages. Copy and clock remain available under their existing message chrome, and the Host's completed-turn fork semantics remain unchanged.
+
+This narrows the message eligibility established by the earlier [Web session fork action decision](../feature/2026-07-27-web-session-fork-actions.md). Session-row forking still selects the latest completed turn, and eligible message actions still pass their event seq through the shared client runtime operation.
+
+## Alternatives considered
+
+**Cut the event log at the clicked assistant message.** Rejected because an assistant message can sit inside an open step and can contain tool calls whose results occur later. A raw prefix at that seq is not a balanced turn and may not be a valid provider transcript.
+
+**Infer completion from `running` or the next user message.** Rejected because retry and steering turns need not align with the next visible user bubble, and a paged window may omit that later bubble. The durable `turn/end` event is the authoritative completion fact.
+
+**Hide branch from every interrupted turn.** Rejected because an aborted turn is durably closed and its final interrupted text can be the true transcript tail. Eligibility depends on the completed boundary and node order, not the outcome kind.
+
+## Consequences
+
+A branch icon now denotes the same completed-turn boundary that the Host will copy. In the reported response → tool → interrupted Think shape, the response keeps copy and clock but no longer advertises branch. This change deliberately does not provide same-turn transcript editing or a retry-before-turn operation; the Session-row action remains available when a reader wants to copy the latest completed turn in full. Runtime tests pin boundary projection and reference stability, while conversation tests cover normal assistant tails, user-only tails, and suppression by later tool and interrupted reasoning rows.

.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.zh.md

@@ -0,0 +1,27 @@
+# Agent Note: 消息 fork 操作要求消息位于已完成轮次尾部
+
+Status: implemented
+
+[English](2026-08-02-message-fork-actions-require-completed-turn-tail.md) | 中文
+
+## 问题
+
+Web 会话把分支操作挂到每个轮次中最后一个文本非空的 assistant 节点上。如果后面还有工具结果、被中断的推理（reasoning）节点或终态错误，这些行也不会接管操作，因为它们没有内容文本 IconActions。因此，分支图标可能出现在 assistant 响应下方，而同一轮次的更多行仍位于其后。Host 会正确地把该消息锚点扩展到其所在的 `turn/end`，但图标位置使操作看起来像在消息级截断，子会话又会明显继承同轮次的后缀。
+
+## 决策
+
+`ConversationSnapshot.turnEnds` 保留原始事件窗口中的已完成轮次边界。会话视图按各边界遍历 transcript（文本记录）节点，仅当边界的最后一个节点是用户消息或含内容的 assistant 消息时才暴露分支操作。开放轮次没有符合条件的消息；如果后面还有工具结果、只有推理内容的中断、轮次错误或其他 transcript 节点，较早消息上的分支操作就会被抑制。复制和时钟仍可在既有消息 chrome 下使用，Host 按已完成轮次 fork 的语义保持不变。
+
+本决策收紧了较早的 [Web 会话 fork 操作决策](../feature/2026-07-27-web-session-fork-actions.md)所定义的消息资格。Session 行 fork 仍选择最新的已完成轮次；符合条件的消息操作仍通过共享 client 运行时操作传递其事件 seq。
+
+## 考虑过的替代方案
+
+**在点击的 assistant 消息处截断事件日志。** 不予采纳：assistant 消息可能位于尚未结束的步骤内，也可能包含结果随后才出现的工具调用。以该 seq 截取的原始前缀并不是结构完整的轮次，也可能不是有效的提供方 transcript。
+
+**从 `running` 或下一条用户消息推断完成状态。** 不予采纳：重试轮次与 steering（中途引导）轮次不一定和下一个可见用户气泡对齐，分页窗口也可能省略该气泡。持久 `turn/end` 事件才是权威的完成事实。
+
+**对每个被中断轮次隐藏分支。** 不予采纳：已中止的轮次会持久关闭，其最终的中断文本可能正是真正的 transcript 尾部。资格取决于已完成边界与节点顺序，而非结果类别。
+
+## 后果
+
+分支图标现在表示的已完成轮次边界与 Host 实际复制的边界一致。在所报告的「响应 → 工具 → 被中断的 Think」形态中，响应仍保留复制和时钟，但不再显示分支。本变更刻意不提供同轮次 transcript 编辑，也不提供轮次前重试操作；当读者希望完整复制最新的已完成轮次时，仍可使用 Session 行操作。运行时测试固定边界投影和引用稳定性，会话测试则覆盖普通 assistant 尾部、纯用户消息尾部，以及后续工具行和被中断推理行对分支操作的抑制。

.agents/notes/implemented/feature/2026-07-27-web-session-fork-actions.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-web-session-fork-actions.md
-2026-07-27-web-session-fork-actions.md: 58960169a2e499d953840e5769e7689b5cd48047
-2026-07-27-web-session-fork-actions.zh.md: ea2f9030f672f00fb91bce3546836689a7d41004
+2026-07-27-web-session-fork-actions.md: 578e59ec92e003fe5c8cdfe951a595f3a7371ecf
+2026-07-27-web-session-fork-actions.zh.md: d90124f6e6e164b0a1e0fce734f52976630cd848

.agents/notes/implemented/feature/2026-07-27-web-session-fork-actions.md

@@ -10,7 +10,9 @@ The Session store already provides a fork primitive that creates a child session
 
 ## Decision
 
-The Web Session-row menu and message IconActions share the client runtime's `sessions.fork` action. A Session row passes `{ sessionId, increaseTitle: true }`, so it forks at the source session's last completed turn; a user message or settled assistant content message passes `{ sessionId, atSeq: node.seq, increaseTitle: true }`, so it forks at the turn containing that event. Only the client consumes `increaseTitle`: after adding the child session to its local list, the client increments a trailing `(N)` or `（N）` in the source session's persisted title without changing bracket style, appends ` (1)` to an unnumbered title, and skips the rename when no persisted title exists; the Host fork request still contains only `sessionId` and the optional `atSeq`. The caller opens the child only after the rename succeeds; a fork or rename failure leaves the source session and current selection unchanged, while a child created before a rename failure remains in the list.
+The message-eligibility portion of this decision is narrowed by the [completed-turn-tail decision](../bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md); the shared runtime action, injection ownership, title handling, and peer-list decisions remain current.
+
+The Web Session-row menu and message IconActions share the client runtime's `sessions.fork` action. A Session row passes `{ sessionId, increaseTitle: true }`, so it forks at the source session's last completed turn; an eligible completed-turn-tail message passes `{ sessionId, atSeq: node.seq, increaseTitle: true }`, so it forks at the turn ending at that message. Only the client consumes `increaseTitle`: after adding the child session to its local list, the client increments a trailing `(N)` or `（N）` in the source session's persisted title without changing bracket style, appends ` (1)` to an unnumbered title, and skips the rename when no persisted title exists; the Host fork request still contains only `sessionId` and the optional `atSeq`. The caller opens the child only after the rename succeeds; a fork or rename failure leaves the source session and current selection unchanged, while a child created before a rename failure remains in the list.
 
 `forkAt(seq)` touches the session service only in ui-conversation's apply injection layer; message components report only the event `seq`. Session rows likewise initiate the operation only through ui-workspace's injected callback. Neither presentation package owns session mutation state or duplicates the host's boundary evaluation.
 
@@ -28,6 +30,6 @@ Session lineage is not projected into a list hierarchy. WorkSpace mode displays
 
 ## Consequences
 
-Users can create forks from Session rows, user messages, or settled assistant content messages; all three entry points ultimately use the same runtime/host operation. Message entry points preserve the exact event boundary, while the list entry point preserves the “latest completed turn” shortcut. Successive fork titles increment through `(1)`, `(2)`, and so on instead of repeatedly appending `(1)`; titles with fullwidth parentheses retain that style. Every fork child immediately appears as an ordinary peer row, so the list no longer needs session expansion state, recursive nodes, or twist controls.
+Users can create forks from Session rows or eligible completed-turn-tail messages; both entry points ultimately use the same runtime/host operation. Message entry points preserve the exact event boundary, while the list entry point preserves the “latest completed turn” shortcut. Successive fork titles increment through `(1)`, `(2)`, and so on instead of repeatedly appending `(1)`; titles with fullwidth parentheses retain that style. Every fork child immediately appears as an ordinary peer row, so the list no longer needs session expansion state, recursive nodes, or twist controls.
 
-Fork and child-rename failures stay silent and preserve the source selection, preventing a derivation action from disrupting the current reading position; this tradeoff also means the UI does not yet expose a failure reason or retry entry point. Package tests separately pin the two message `seq` paths, title increments, and the peer-list derivation; `apps/web/tests/message-actions.e2e.ts` exercises assistant-message branching and Session-row menu branching through the assembled application.
+Fork and child-rename failures stay silent and preserve the source selection, preventing a derivation action from disrupting the current reading position; this tradeoff also means the UI does not yet expose a failure reason or retry entry point. Package tests pin eligible message `seq` forwarding, title increments, and the peer-list derivation; `apps/web/tests/message-actions.e2e.ts` exercises assistant-message branching and Session-row menu branching through the assembled application.

.agents/notes/implemented/feature/2026-07-27-web-session-fork-actions.zh.md

@@ -10,7 +10,9 @@ Session store 已提供按完成轮前缀创建子会话的 fork 原语，但 We
 
 ## Decision
 
-Web 的 session 行菜单与消息 IconActions 共用 client runtime 的 `sessions.fork` 操作。Session 行传 `{ sessionId, increaseTitle: true }`，因此在源会话最后一个已完成轮次处分支；用户消息与已定稿 assistant 内容消息传 `{ sessionId, atSeq: node.seq, increaseTitle: true }`，因此在包含该事件的轮次处分支。`increaseTitle` 只由 client 消费：子会话进入本地列表后，client 把源会话持久化标题尾部的 `(N)` 或 `（N）` 递增并保留括号样式，无编号时追加 ` (1)`，没有持久化标题时不改名；Host fork 请求仍只有 `sessionId` 与可选的 `atSeq`。改名成功后调用方才打开子会话；fork 或改名失败时保持源会话与当前选择不变，改名失败时已创建的子会话仍留在列表中。
+本决策中的消息资格部分由[已完成轮次尾部决策](../bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md)收紧；共享运行时操作、注入归属、标题处理和同级列表决策仍然有效。
+
+Web 的 session 行菜单与消息 IconActions 共用 client runtime 的 `sessions.fork` 操作。Session 行传 `{ sessionId, increaseTitle: true }`，因此在源会话最后一个已完成轮次处分支；符合条件且位于已完成轮次尾部的消息传 `{ sessionId, atSeq: node.seq, increaseTitle: true }`，因此在以该消息结束的轮次处分支。`increaseTitle` 只由 client 消费：子会话进入本地列表后，client 把源会话持久化标题尾部的 `(N)` 或 `（N）` 递增并保留括号样式，无编号时追加 ` (1)`，没有持久化标题时不改名；Host fork 请求仍只有 `sessionId` 与可选的 `atSeq`。改名成功后调用方才打开子会话；fork 或改名失败时保持源会话与当前选择不变，改名失败时已创建的子会话仍留在列表中。
 
 `forkAt(seq)` 只在 ui-conversation 的 apply 注入层接触 session 服务，消息组件只回传事件 `seq`。Session 行同理只通过 ui-workspace 的注入回调发起操作；两个呈现包都不持有 session mutation 状态，也不复制 host 的边界求值。
 
@@ -28,6 +30,6 @@ Session lineage 不投影成列表层级。WorkSpace 模式按 `WorkspaceView.se
 
 ## Consequences
 
-用户可从 session 行、用户消息或已定稿 assistant 内容消息创建分支，三处最终走同一个 runtime/host 操作；消息点位保留精确事件边界，列表点位保留「最新完成轮」快捷语义。连续 fork 的标题按 `(1)`、`(2)` 递增，而不是重复追加 `(1)`；全角括号标题保持全角样式。所有 fork 子会话立即作为普通同级行出现，列表不再需要 session 展开状态、递归节点或 twist 控件。
+用户可从 session 行或符合条件的已完成轮次尾部消息创建分支，两处最终走同一个 runtime/host 操作；消息点位保留精确事件边界，列表点位保留「最新完成轮」快捷语义。连续 fork 的标题按 `(1)`、`(2)` 递增，而不是重复追加 `(1)`；全角括号标题保持全角样式。所有 fork 子会话立即作为普通同级行出现，列表不再需要 session 展开状态、递归节点或 twist 控件。
 
-Fork 与子会话改名失败都保持静默并保留源选择，避免一个派生操作破坏当前阅读位置；该取舍也意味着 UI 暂不提供失败原因或重试入口。Package tests 分别钉住两种消息 `seq`、标题递增与同级列表派生，`apps/web/tests/message-actions.e2e.ts` 通过装配后的应用执行 assistant 消息分支与 session 行菜单分支。
+Fork 与子会话改名失败都保持静默并保留源选择，避免一个派生操作破坏当前阅读位置；该取舍也意味着 UI 暂不提供失败原因或重试入口。包级测试固定符合条件的消息 `seq` 转发、标题递增与同级列表派生，`apps/web/tests/message-actions.e2e.ts` 通过装配后的应用执行 assistant 消息分支与 session 行菜单分支。

apps/web/tests/message-actions.e2e.ts

@@ -1,7 +1,7 @@
-// Web e2e scenario: message IconActions + clocks. Cold-seeds the seeded-history
-// fixture (zero model calls) and pins the settled conversation aria after the
-// user/assistant footers are focus-revealed — the surface package jsdom tests
-// cannot substitute for (docs/testing.md snapshot rule).
+// Web e2e scenario: message IconActions + clocks. Cold-seeds a deterministic
+// completed-turn-tail fork case (zero model calls) and pins the settled
+// conversation aria after the footers are focus-revealed — the surface package
+// jsdom tests cannot substitute for (docs/testing.md snapshot rule).
 import { mkdir, readFile, writeFile } from 'node:fs/promises'
 import { join } from 'node:path'
 import { fileURLToPath } from 'node:url'
@@ -25,6 +25,48 @@ const MODE = webSnapshotMode()
 const SEED_ID = 'message-actions-web-e2e'
 
 const PROMPT = 'Use the read tool twice in one assistant message: read a.txt and b.txt. Then reply with the single word DONE and stop.'
+const MID_TURN_TEXT = 'I will read both files before answering.'
+const SECOND_PROMPT = 'Now give the final answer.'
+
+/**
+ * Adapt the borrowed recording into response -> tools -> interrupted Think,
+ * followed by one ordinary completed response. The first response keeps
+ * copy/clock but is not a legal branch point; the second is the real turn tail.
+ * @param raw - Recorded seeded-history JSONL.
+ * @returns A contiguous, closed two-turn fixture.
+ */
+function completedTailFixture(raw: string): string {
+  const kept: string[] = []
+  for (const line of raw.trimEnd().split('\n')) {
+    const row = JSON.parse(line) as {
+      type: string
+      seq?: number
+      seq0?: number
+      data?: { content?: unknown[] }
+    }
+    const firstSeq = row.seq ?? row.seq0
+    if (firstSeq !== undefined && firstSeq >= 101) break
+    if (row.type === 'assistant/message' && row.seq === 64) {
+      const content = row.data?.content
+      if (!Array.isArray(content)) throw new Error('borrowed step-one assistant message has no content')
+      content.splice(1, 0, { type: 'text', text: MID_TURN_TEXT })
+      kept.push(JSON.stringify(row))
+    } else {
+      kept.push(line)
+    }
+  }
+  const tail = [
+    { type: 'step/end', seq: 101, time: 1784974102749, data: { turn: 1, step: 2 } },
+    { type: 'turn/end', seq: 102, time: 1784974102750, data: { turn: 1, reason: { kind: 'aborted' } } },
+    { type: 'turn/start', seq: 103, time: 1784974103000, data: { turn: 2, trigger: { kind: 'message', source: { kind: 'user', rpcId: '{{rpcId}}' } } } },
+    { type: 'user/message', seq: 104, time: 1784974103001, data: { content: [{ type: 'text', text: SECOND_PROMPT }], source: { kind: 'user', rpcId: '{{rpcId}}' } }, surfaceOp: 'append' },
+    { type: 'step/start', seq: 105, time: 1784974103002, data: { turn: 2, step: 1 } },
+    { type: 'assistant/message', seq: 106, time: 1784974103003, data: { turn: 2, step: 1, content: [{ type: 'text', text: 'DONE' }], provenance: { provider: 'deepseek-official', model: 'deepseek-v4-flash' } }, sourceEventSeqs: [], surfaceOp: 'append' },
+    { type: 'step/end', seq: 107, time: 1784974103004, data: { turn: 2, step: 1 } },
+    { type: 'turn/end', seq: 108, time: 1784974103005, data: { turn: 2, reason: { kind: 'completed' } } },
+  ]
+  return `${[...kept, ...tail.map(row => JSON.stringify(row))].join('\n')}\n`
+}
 
 describe('web e2e: message IconActions and clocks on settled history', () => {
   let scaffold: WebScaffold
@@ -38,8 +80,8 @@ describe('web e2e: message IconActions and clocks on settled history', () => {
     await mkdir(sessionCwd, { recursive: true })
     await writeFile(join(sessionCwd, 'a.txt'), 'alpha\n')
     await writeFile(join(sessionCwd, 'b.txt'), 'beta\n')
-    const raw = await readFile(SEED, 'utf8')
-    expect(fixtureUserPrompts(raw), 'borrowed seed must carry the drive prompt').toEqual([PROMPT])
+    const raw = completedTailFixture(await readFile(SEED, 'utf8'))
+    expect(fixtureUserPrompts(raw), 'adapted seed must carry both prompts').toEqual([PROMPT, SECOND_PROMPT])
     await seedSession(scaffold, raw, SEED_ID)
     browser = await chromium.launch()
     page = await newEnglishPage(browser)
@@ -53,7 +95,7 @@ describe('web e2e: message IconActions and clocks on settled history', () => {
     await scaffold?.close()
   })
 
-  it.skipIf(MODE === 'record')('lists the seeded session and reveals user/assistant IconActions', async () => {
+  it.skipIf(MODE === 'record')('shows branch only on the completed transcript tail', async () => {
     onTestFailed(() => saveFailureShot(page, 'web-e2e-message-actions'))
     const groupRow = page.locator('[role="treeitem"]').first()
     await groupRow.waitFor({ timeout: 15_000 })
@@ -61,16 +103,17 @@ describe('web e2e: message IconActions and clocks on settled history', () => {
     const sessionRow = page.locator('[role="treeitem"]').nth(1)
     await sessionRow.waitFor({ timeout: 10_000 })
     await sessionRow.click()
+    await expect.poll(() => page.getByText(MID_TURN_TEXT, { exact: true }).count(), { timeout: 15_000 }).toBe(1)
     await expect.poll(() => page.getByText('DONE', { exact: true }).count(), { timeout: 15_000 }).toBe(1)
 
     // Focus-reveal the footers (hover:hover keeps them opacity-hidden until
-    // hover/focus-within). User and each turn's last content assistant both
-    // have copy + branch.
+    // hover/focus-within). All message rows keep copy, but only the final
+    // assistant at a completed transcript tail has branch.
     const copyButtons = page.getByRole('button', { name: 'Copy' })
-    await expect.poll(() => copyButtons.count(), { timeout: 10_000 }).toBeGreaterThanOrEqual(2)
+    await expect.poll(() => copyButtons.count(), { timeout: 10_000 }).toBeGreaterThanOrEqual(4)
     await copyButtons.first().focus()
     await expect.poll(() => page.getByRole('button', { name: 'Branch into a new conversation' }).count(), { timeout: 5_000 })
-      .toBeGreaterThanOrEqual(2)
+      .toBe(1)
     await expect.poll(() => page.getByRole('button', { name: 'Edit' }).count(), { timeout: 5_000 }).toBe(0)
   }, 60_000)
 
@@ -89,8 +132,7 @@ describe('web e2e: message IconActions and clocks on settled history', () => {
 
   it.skipIf(MODE === 'record')('forks through the settled-message and session-row actions', async () => {
     onTestFailed(() => saveFailureShot(page, 'web-e2e-message-fork'))
-    // Exercise the assistant action specifically; package coverage pins the
-    // user action separately at its own event seq.
+    // The sole message action belongs to the completed second-turn assistant.
     await page.getByRole('button', { name: 'Branch into a new conversation' }).last().click()
     await expect.poll(
       () => scaffold.ctx.agents.list().find(agent => agent.session.header.parentSession === SessionId(SEED_ID)),

apps/web/tests/snapshots/message-actions/ui.expected.md

@@ -8,12 +8,14 @@
 - button "Copy":
   - img
 - tooltip "Copy"
-- button "Branch into a new conversation":
-  - img
 - button "Think The user wants me to read a.txt and b.txt, then reply with \"DONE\". Let me do both reads in parallel.":
   - img
   - img
   - text: Think The user wants me to read a.txt and b.txt, then reply with "DONE". Let me do both reads in parallel.
+- paragraph: I will read both files before answering.
+- button "Copy":
+  - img
+- text: 7/25 {{clock}}
 - button "Read a.txt":
   - img
   - img
@@ -28,6 +30,9 @@
   - img
   - img
   - text: Think Both files have been read. a.txt contains "alpha" and b.txt contains "beta". I'll now reply with DONE as instructed.
+- text: Stopped Now give the final answer. 7/25 {{clock}}
+- button "Copy":
+  - img
 - paragraph: DONE
 - button "Copy":
   - img
@@ -42,4 +47,4 @@
   - text: deepseek-v4-flash
   - img
 - button "Send message" [disabled]
-- text: 1 turns · 2 steps Tool call {{duration}} Cache hit 98% Input 15.8K tok · Output 135 tok
+- text: 2 turns · 3 steps Tool call {{duration}} Cache hit 98% Input 7.8K tok · Output 103 tok

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: f956be22384a42e9ed30e8aa5f25fe8173cc9f9c
-README.zh.md: 49449c51d89d957b5bd39798c9167607f72a5c3a
+README.md: 91844016ce58e172161343022a1824be91b7ff2f
+README.zh.md: 0ea77774c6181858748a8adbc737de7e213be31b

packages/client/runtime/README.md

@@ -26,7 +26,7 @@ SlotsService gives the renderer separate bare observables for `useSessions` and
 
 ## The human transcript
 
-`ConversationSnapshot.nodes` is the human transcript, not the model surface. `TranscriptAdapter` projects the raw window in log order — every append-origin surface event (`isAppendSurfaceEvent`) at its own log position, plus one `CompactionSummaryNode` marker per landed compaction checkpoint — and never consults surface order. A landed compaction therefore keeps the conversation it shadowed on the model side: the marker reports where the model stopped seeing that history instead of erasing it. Model-only replacement copies stay out: a pruned `tool/result` and a regenerated `assistant/message` rewrite one node for the model and mark no boundary. A checkpoint is a `user/message` carrying the compaction seam's plugin source that **replaced** a surface range; an appending plugin-sourced `user/message` is injected context, not a compaction. The adapter's plugin literal is pinned to the seam's own declaration by a type-only import of the cordis-free [`dsh-compact/checkpoint`](../../compact/compact/README.md) leaf, so renaming it there fails `tsc` here; a **value** import of the package would fail the client purity gate, and the package **root** is unreachable even as a type (it reaches `dsh-session`'s root, whose `Context` merge collides the host `sessions` with this program's). `tests/compact-checkpoint-pin.spec.ts` covers the same drift behaviorally.
+`ConversationSnapshot.nodes` is the human transcript, not the model surface. `TranscriptAdapter` projects the raw window in log order — every append-origin surface event (`isAppendSurfaceEvent`) at its own log position, plus one `CompactionSummaryNode` marker per landed compaction checkpoint — and never consults surface order. `ConversationSnapshot.turnEnds` maps each completed turn in that window to its `turn/end` seq, retaining turn completion independently from the transcript so presentation can require a real boundary before exposing an action. A landed compaction therefore keeps the conversation it shadowed on the model side: the marker reports where the model stopped seeing that history instead of erasing it. Model-only replacement copies stay out: a pruned `tool/result` and a regenerated `assistant/message` rewrite one node for the model and mark no boundary. A checkpoint is a `user/message` carrying the compaction seam's plugin source that **replaced** a surface range; an appending plugin-sourced `user/message` is injected context, not a compaction. The adapter's plugin literal is pinned to the seam's own declaration by a type-only import of the cordis-free [`dsh-compact/checkpoint`](../../compact/compact/README.md) leaf, so renaming it there fails `tsc` here; a **value** import of the package would fail the client purity gate, and the package **root** is unreachable even as a type (it reaches `dsh-session`'s root, whose `Context` merge collides the host `sessions` with this program's). `tests/compact-checkpoint-pin.spec.ts` covers the same drift behaviorally.
 
 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.
 

packages/client/runtime/README.zh.md

@@ -26,7 +26,7 @@ SlotsService 分别为 renderer 提供 `useSessions` 与 `useWorkspaces` 的裸
 
 ## 面向人的 transcript（文本记录）
 
-`ConversationSnapshot.nodes` 是面向人的 transcript，不是模型 surface。`TranscriptAdapter` 按日志顺序投影原始窗口——每个 append 来源的 surface 事件（`isAppendSurfaceEvent`）落在它自己的日志位置上，外加每次落地的压缩（compaction）检查点贡献一个 `CompactionSummaryNode` 标记——且从不查询 surface 顺序。于是一次落地的压缩会保留它在模型侧遮蔽掉的对话：标记报告模型从哪里开始看不见那段历史，而不是把它抹掉。仅模型可见的 replacement 副本不进入记录：被裁剪的 `tool/result` 和重新生成的 `assistant/message` 只为模型重写一个节点，不标记任何边界。检查点是携带压缩 seam 插件来源、且**替换**了一段 surface 范围的 `user/message`；一条 append 的插件来源 `user/message` 是注入上下文，不是压缩。适配器的插件字面量通过对无 cordis 的 [`dsh-compact/checkpoint`](../../compact/compact/README.md) 叶子做仅类型导入，钉在压缩 seam 自己的声明上：在那里改名会让此处 `tsc` 失败；而对该包（package）做**值**导入会被客户端纯度门禁拒绝，包的**根**即便作为类型也无法到达（它会到达 `dsh-session` 的根，其 `Context` 合并会让 host 的 `sessions` 与本程序的冲突）。`tests/compact-checkpoint-pin.spec.ts` 从行为侧覆盖同一漂移。
+`ConversationSnapshot.nodes` 是面向人的 transcript，不是模型 surface。`TranscriptAdapter` 按日志顺序投影原始窗口——每个 append 来源的 surface 事件（`isAppendSurfaceEvent`）落在它自己的日志位置上，外加每次落地的压缩（compaction）检查点贡献一个 `CompactionSummaryNode` 标记——且从不查询 surface 顺序。`ConversationSnapshot.turnEnds` 把该窗口中的每个已完成轮次映射到其 `turn/end` seq；它独立于 transcript 保留轮次完成状态，使呈现层能够在暴露操作前要求存在真实边界。于是一次落地的压缩会保留它在模型侧遮蔽掉的对话：标记报告模型从哪里开始看不见那段历史，而不是把它抹掉。仅模型可见的 replacement 副本不进入记录：被裁剪的 `tool/result` 和重新生成的 `assistant/message` 只为模型重写一个节点，不标记任何边界。检查点是携带压缩 seam 插件来源、且**替换**了一段 surface 范围的 `user/message`；一条 append 的插件来源 `user/message` 是注入上下文，不是压缩。适配器的插件字面量通过对无 cordis 的 [`dsh-compact/checkpoint`](../../compact/compact/README.md) 叶子做仅类型导入，钉在压缩 seam 自己的声明上：在那里改名会让此处 `tsc` 失败；而对该包（package）做**值**导入会被客户端纯度门禁拒绝，包的**根**即便作为类型也无法到达（它会到达 `dsh-session` 的根，其 `Context` 合并会让 host 的 `sessions` 与本程序的冲突）。`tests/compact-checkpoint-pin.spec.ts` 从行为侧覆盖同一漂移。
 
 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。
 

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

@@ -322,6 +322,8 @@ export interface ConversationSnapshot {
   sessionId: SessionId
   /** Human transcript plus retry notices and interrupted-turn terminal nodes in event order. */
   nodes: readonly ConversationNode[]
+  /** In-window completed turn number -> its `turn/end` event seq. */
+  turnEnds: ReadonlyMap<number, number>
   partial: PartialAssistant | null
   runningCalls: readonly RunningToolCall[]
   /**

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

@@ -113,6 +113,11 @@ export class Session implements SessionFace {
   private pendingCache: { rev: number; value: PendingInteraction[] } | null = null
   private derivedRev = 0
   private nodesCache: { projected: readonly ConversationNode[]; derivedRev: number; value: readonly ConversationNode[] } | null = null
+  /** Completed turn boundaries retained from the raw window so presentation
+   *  actions never infer a safe fork point from transcript content alone. */
+  private turnEnds = new Map<number, number>()
+  private turnEndsRev = 0
+  private turnEndsCache: { rev: number; value: ReadonlyMap<number, number> } | null = null
   /** Authoritative stream-only inbox snapshot; pending work never hits history. */
   private queued: QueuedMessage[] = []
   private queueRev = 0
@@ -807,6 +812,8 @@ export class Session implements SessionFace {
         return
       }
       case 'turn/end': {
+        this.turnEnds.set(event.data.turn, event.seq)
+        this.turnEndsRev++
         if (event.data.reason.kind === 'aborted' || event.data.reason.kind === 'disposed') {
           this.settleScheduledRetry('cancelled', event.data.turn)
         }
@@ -897,6 +904,8 @@ export class Session implements SessionFace {
     this.callsRev++
     this.derivedNodes = []
     this.derivedRev++
+    this.turnEnds = new Map()
+    this.turnEndsRev++
     this.codeDispatches = new Map()
     this.dispatchesRev++
     for (let i = 0; i < this.events.length; i++) {
@@ -928,6 +937,9 @@ export class Session implements SessionFace {
     if (this.callsCache === null || this.callsCache.rev !== this.callsRev) {
       this.callsCache = { rev: this.callsRev, value: [...this.openCalls.values()] }
     }
+    if (this.turnEndsCache === null || this.turnEndsCache.rev !== this.turnEndsRev) {
+      this.turnEndsCache = { rev: this.turnEndsRev, value: new Map(this.turnEnds) }
+    }
     if (this.pendingCache === null || this.pendingCache.rev !== this.pendingRev) {
       this.pendingCache = { rev: this.pendingRev, value: [...this.pending.values()] }
     }
@@ -941,6 +953,7 @@ export class Session implements SessionFace {
     return {
       sessionId: this.sessionId,
       nodes,
+      turnEnds: this.turnEndsCache.value,
       partial,
       runningCalls: this.callsCache.value,
       pending: this.pendingCache.value,

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

@@ -426,6 +426,7 @@ describe('live event path', () => {
     feed(ev.turnEnd(10, 1, 'aborted')) // no assistant/message ever arrives
     const snapshot = session.getSnapshot()
     expect(snapshot.partial).toBeNull()
+    expect(snapshot.turnEnds.get(1)).toBe(10)
     const frozen = snapshot.nodes.at(-1)
     expect(frozen).toMatchObject({ kind: 'assistant', interrupted: true, blocks: [{ kind: 'text', text: '说到一半' }] })
     // Ordered inside the flow: after the user message (seq 7), before any later turn.
@@ -1215,6 +1216,7 @@ describe('reference stability (the memo contract)', () => {
     expect(after).not.toBe(before)
     expect(after.runningCalls).toBe(before.runningCalls)
     expect(after.pending).toBe(before.pending)
+    expect(after.turnEnds).toBe(before.turnEnds)
     // And a mutation on the tracked domain swaps that array.
     feed(ev.toolResult(11, 1, 'c1', 'ECHO'))
     const resolved = session.getSnapshot()

packages/client/test-runtime/src/fixtures.ts

@@ -46,6 +46,7 @@ export function conversationSnapshot(sessionId: SessionId): ConversationSnapshot
   return {
     sessionId,
     nodes: [],
+    turnEnds: new Map(),
     partial: null,
     runningCalls: [],
     codeDispatches: new Map(),

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: 65e4542e622713e2cd120906926fc0601ef280bd
-README.zh.md: 4828ad214bb5ec0a9921307dd01dcc282910b330
+README.md: e351880066ac00677ab9d22c96ea75e2244a2ecf
+README.zh.md: f7c7e3ab6d341f9b3d0866951ab80a18d5968044

packages/client/ui-conversation/README.md

@@ -57,8 +57,8 @@ None; this package neither assembles nor sends a provider request.
 - **Compaction markers show no scale** — the row does not yet report how many messages or which range the checkpoint replaced.
 - **Stats-line durations cover the in-window flow only** — LLM and tool wall times fold the snapshot's assistant `timing` and tool call/result pairs, so nodes outside the loaded event window (older history) are not counted.
 - **Details panel is the minimal form and currently has no entry point** — selected call args/result raw display; the Input/Output/Metadata switch, Prev/Next stepping, and See-in-trajectory deep link are deferred. Tool rows stopped being details-panel click targets and nothing replaced that gesture, so `ChatViewInjected.openDetails` is implemented but uncalled and the panel (including its terminal card) is unreachable in the assembled application; its rendering stays covered by mounting it with a selection directly.
-- **Assistant per-message paging is a reserved slot** — drawn in the design, not implemented. The finalized content IconActions row (copy / branch / clock) ships under the last content-text assistant of each turn only; mid-turn narration and Think-only nodes stay chrome-free. Branch forks through the turn containing that message, increments the inherited title on the client, and then opens the child, while a fork or rename failure leaves the source selected.
-- **Sent user messages cannot be edited** — the user bubble's IconActions row carries clock / copy / branch only, and branching from the message is the nearest gesture. The control returns with the capability behind it: a client mutation over a settled user message, plus the host behavior for the turn that already consumed it ([decision](../../../.agents/notes/implemented/simplification/2026-07-31-drop-user-message-edit-stub.md)).
+- **Assistant per-message paging is a reserved slot** — drawn in the design, not implemented. The finalized content IconActions row (copy / clock, plus branch when eligible) ships under the last content-text assistant of each turn only; mid-turn narration and Think-only nodes stay chrome-free. Branch appears only when that message is also the last transcript node of a completed turn, then forks through that turn, increments the inherited title on the client, and opens the child; a fork or rename failure leaves the source selected ([decision](../../../.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md)).
+- **Sent user messages cannot be edited** — user bubbles retain clock and copy, while branch appears only for a completed turn whose transcript ends at that user message. Editing returns with the capability behind it: a client mutation over a settled user message, plus the host behavior for the turn that already consumed it ([decision](../../../.agents/notes/implemented/simplification/2026-07-31-drop-user-message-edit-stub.md)).
 - **The sparkle icon for the others tool row is a hand-drawn approximation** — the design glyph's vector geometry is not exportable locally; promotion into ui-primitives waits on an exact export.
 - **The approval panel's "Always allow this type" is deferred** — durable grants need a grant-storage design; only allow-once/reject answer today.
 - **TodoPanel truncates long item text to one ellipsized line** — the figma strip has no wrap or expand affordance; full text is not readable inline.

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

@@ -57,8 +57,8 @@ todo 两个面就是在该形状上的两个注册项，都是普通注册方插
 - **压缩标记不显示规模**：该行尚不报告检查点替换了多少条消息或哪段范围。
 - **统计行的耗时只覆盖窗口内消息流**：LLM 与工具墙钟时间由快照的 assistant `timing` 与工具 call/result 配对折算，落在已加载事件窗口之外的节点（更早的历史）不计入。
 - **详情面板是最小形态，且当前没有入口**：以原始形式显示已选择调用的参数／结果；Input/Output/Metadata 切换、Prev/Next 步进与 See-in-trajectory 深链接暂缓实现。工具行已不再是详情面板的点击目标，且没有任何手势接替它，因此 `ChatViewInjected.openDetails` 虽已实现却无人调用，该面板（含其终端卡片）在组装后的应用中不可达；其渲染仍由直接以选中态挂载它来覆盖。
-- **assistant 逐消息分页是预留 slot**：设计中已有图稿，尚未实现。已定稿的内容 IconActions 行（复制／分支／时钟）只挂在每个轮次中最后一条带 text 内容的 assistant 下；轮次中间的叙述与纯 Think 节点不带 chrome。分支会 fork 到包含该消息的轮次末尾，在 client 端递增继承标题后打开子会话，而 fork 或改名失败时源会话保持选中。
-- **已发送的 user 消息无法编辑**：user 气泡的 IconActions 行只有时钟／复制／分支，从该消息分支是最接近的手势。该控件要与其背后的能力一起回归：既需要针对已定稿 user 消息的 client 变更，也需要 host 侧对已经消费过它的轮次给出行为（[决策](../../../.agents/notes/implemented/simplification/2026-07-31-drop-user-message-edit-stub.md)）。
+- **assistant 逐消息分页是预留 slot**：设计中已有图稿，尚未实现。已定稿的内容 IconActions 行（复制／时钟，符合条件时再显示分支）只挂在每个轮次中最后一条带 text 内容的 assistant 下；轮次中间的叙述与纯 Think 节点不带 chrome。只有当该消息同时也是已完成轮次的最后一个 transcript 节点时才显示分支；随后 fork 到该轮次末尾，在 client 端递增继承标题并打开子会话；fork 或改名失败时源会话保持选中（[决策](../../../.agents/notes/implemented/bug-fix/2026-08-02-message-fork-actions-require-completed-turn-tail.md)）。
+- **已发送的 user 消息无法编辑**：user 气泡保留时钟和复制；仅当已完成轮次的 transcript 结束于该 user 消息时才显示分支。编辑功能要与其背后的能力一起回归：既需要针对已定稿 user 消息的 client 变更，也需要 host 侧对已经消费过它的轮次给出行为（[决策](../../../.agents/notes/implemented/simplification/2026-07-31-drop-user-message-edit-stub.md)）。
 - **others 工具行的闪光图标是手绘近似版本**：无法在本地导出设计字形的矢量几何；等到存在精确导出后再将其提升到 ui-primitives。
 - **审批面板的「始终允许此类」暂缓**：持久授权需要授权存储设计；今天只能回答允许一次／拒绝。
 - **TodoPanel 将过长条目截成单行省略号**：figma 条没有换行或展开入口，完整文本无法在行内读完。

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

@@ -4,9 +4,10 @@
 // view groups them into tool rows through its keyed toolview slot (figma
 // step-summary flow). Shared by finalized nodes and the streaming partial;
 // the turn-level loading dots live in the chat view's tail, not here.
-// Finalized turn-tail content (text) nodes append IconActions once streaming
-// ends (`time` is omitted for mid-turn narration); Think / tool-head-only
-// nodes stay chrome-free.
+// Finalized content (text) nodes append IconActions once streaming ends
+// (`time` is omitted for mid-turn narration); their branch action is present
+// only when the node is also the completed turn's transcript tail. Think /
+// tool-head-only nodes stay chrome-free.
 
 import { memo, useMemo } from 'react'
 import type { AssistantBlock } from '@deepseek-ai/dsh-client-runtime/client'
@@ -28,7 +29,7 @@ export interface AssistantMarkdownProps {
   time?: number | undefined
   /** Event sequence used as the fork boundary; omitted while streaming. */
   seq?: number | undefined
-  /** Fork the session through the turn containing this finalized message. */
+  /** Fork the session through this finalized message's completed turn. */
   onFork?: ((seq: number) => void) | undefined
   /** The owning view's locale seat, passed down as a plain prop. */
   t: ChatViewSlotProps['t']

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

@@ -30,7 +30,7 @@ import type {
 import type { SnapshotSelectorHook } from '@deepseek-ai/dsh-client-ui-slots'
 import { IconChevronDownOutline14 } from '@deepseek-ai/dsh-client-ui-primitives'
 import type { ChatViewSlotProps } from '../contract/slots.ts'
-import { assistantActionsSeqs, deriveChatFlow, type ChatFlowItem } from './chat-flow.ts'
+import { assistantActionsSeqs, deriveChatFlow, messageBranchSeqs, type ChatFlowItem } from './chat-flow.ts'
 import { AssistantMarkdown } from './AssistantMarkdown.tsx'
 import { GenericCommandCard } from './GenericCommandCard.tsx'
 import { GenericToolCard } from './GenericToolCard.tsx'
@@ -236,6 +236,7 @@ export function ChatView({
   useSession, useSessions, useStore, renderSlot, sessionId, openFile, loadOlder, inspectCall, chatScroll, forkAt, t,
 }: ChatViewSlotProps) {
   const nodes = useSession(s => s.nodes)
+  const turnEnds = useSession(s => s.turnEnds)
   // Workspace root off the session list row: path summaries display relative to it.
   const cwd = useSessions(s => s.byId[sessionId]?.cwd)
   const running = useSession(s => s.running)
@@ -252,6 +253,7 @@ export function ChatView({
   // Only the last content assistant of each turn owns IconActions; mid-turn
   // text (before tools) omits `time` so AssistantMarkdown stays chrome-free.
   const actionSeqs = useMemo(() => assistantActionsSeqs(nodes), [nodes])
+  const branchSeqs = useMemo(() => messageBranchSeqs(nodes, turnEnds), [nodes, turnEnds])
 
   const listRef = useRef<HTMLDivElement | null>(null)
   const atBottomRef = useRef(true)
@@ -401,7 +403,7 @@ export function ChatView({
           interrupted={node.interrupted}
           time={actionSeqs.has(node.seq) ? node.time : undefined}
           seq={node.seq}
-          onFork={forkAt}
+          onFork={branchSeqs.has(node.seq) ? forkAt : undefined}
           t={t}
         />
       )
@@ -416,7 +418,7 @@ export function ChatView({
         key={item.key}
         node={node}
         retryActive={node.kind === 'model-retry' && node.seq === activeRetry}
-        onFork={forkAt}
+        {...branchSeqs.has(node.seq) ? { onFork: forkAt } : {}}
         t={t}
       />
     )

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

@@ -17,7 +17,7 @@ export interface MessageIconActionsProps {
   time: number
   /** Clock before icons (user) or after (assistant). */
   clock: 'start' | 'end'
-  /** Fork the session at this message. */
+  /** Fork the session at this message; omission hides the branch action. */
   onBranch?: (() => void) | undefined
   /** Parent layout class composed onto the actions row. */
   className?: string | undefined
@@ -50,11 +50,13 @@ export function MessageIconActions({
           <IconCopyOutline16 />
         </button>
       </Tooltip>
-      <Tooltip label={t('message.branch')} side="bottom">
-        <button type="button" className={css.action} aria-label={t('message.branch')} onClick={onBranch}>
-          <IconBranchOutline16 />
-        </button>
-      </Tooltip>
+      {onBranch !== undefined && (
+        <Tooltip label={t('message.branch')} side="bottom">
+          <button type="button" className={css.action} aria-label={t('message.branch')} onClick={onBranch}>
+            <IconBranchOutline16 />
+          </button>
+        </Tooltip>
+      )}
       {clock === 'end' ? clockEl : null}
     </div>
   )

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

@@ -26,7 +26,7 @@ export interface MessageItemProps {
     | TurnErrorNode
     | UnknownSurfaceNode
   retryActive?: boolean
-  /** Fork the session through the turn containing this message (user-bubble branch action). */
+  /** Fork through this message's completed turn when it is the transcript tail. */
   onFork?: (seq: number) => void
   /** The owning view's locale seat, passed down as a plain prop. */
   t: ChatViewSlotProps['t']

packages/client/ui-conversation/src/client/chat/chat-flow.ts

@@ -5,8 +5,8 @@
  * reuse the first notice's row while projecting the latest retry turn.
  * Item identity keys are stable across snapshots so the list parent can
  * subscribe to keys only while rows subscribe to content. IconActions ownership
- * (last content assistant per turn) is derived here too so ChatView and the
- * flow share one gate.
+ * and completed-turn branch points are derived here too so ChatView and the
+ * flow share their gates.
  */
 import type {
   AssistantBlock, ConversationNode, ToolResultNode,
@@ -47,6 +47,38 @@ export function assistantActionsSeqs(nodes: readonly ConversationNode[]): Readon
   return new Set(lastByTurn.values())
 }
 
+/**
+ * Seq set of message rows that may fork: the last transcript node of a
+ * completed turn, when that node owns message chrome. A later tool, reasoning,
+ * error, or other transcript node suppresses the earlier message's branch
+ * action even though the Host would include the whole turn.
+ * @param nodes - snapshot nodes in event order.
+ * @param turnEnds - completed turn boundaries retained from the event window.
+ * @returns Message seq values whose visible position matches the fork boundary.
+ */
+export function messageBranchSeqs(
+  nodes: readonly ConversationNode[],
+  turnEnds: ReadonlyMap<number, number>,
+): ReadonlySet<number> {
+  const result = new Set<number>()
+  const boundaries = [...turnEnds].sort((a, b) => a[1] - b[1])
+  let nodeIndex = 0
+  for (const [turn, endSeq] of boundaries) {
+    let tail: ConversationNode | undefined
+    while (nodeIndex < nodes.length) {
+      const candidate = nodes[nodeIndex]
+      if (candidate === undefined || candidate.seq > endSeq) break
+      tail = candidate
+      nodeIndex++
+    }
+    if (tail?.kind === 'user'
+      || (tail?.kind === 'assistant' && tail.turn === turn && hasContentText(tail.blocks))) {
+      result.add(tail.seq)
+    }
+  }
+  return result
+}
+
 /**
  * Group finalized nodes into the step-summary flow.
  * @param nodes - snapshot nodes in human-transcript and durable-notice order.

packages/client/ui-conversation/src/client/contract/slots.ts

@@ -454,7 +454,7 @@ export interface ChatViewInjected {
     /** Last recorded offset, or null when pinned or never recorded. */
     read: () => number | null
   }
-  /** Fork the session through the turn containing the message at `seq`, then open the child. */
+  /** Fork through the completed turn ending at the eligible message `seq`, then open the child. */
   forkAt: (seq: number) => void
 }
 

packages/client/ui-conversation/tests/chat-branch-tails.spec.tsx

@@ -36,12 +36,14 @@ describe('MessageItem arms', () => {
     // Same-day clock: construct "today at 14:24" so the label stays `HH:mm`.
     const now = new Date()
     const time = new Date(now.getFullYear(), now.getMonth(), now.getDate(), 14, 24).getTime()
+    const onFork = vi.fn()
     render(
       <MessageItem t={t} node={{
         kind: 'user', seq: 1, time,
         content: [{ type: 'text', text: 'hello bubble' }] as never,
         source: null,
       }}
+      onFork={onFork}
       />,
     )
     expect(screen.getByText('14:24')).toBeTruthy()
@@ -50,6 +52,8 @@ describe('MessageItem arms', () => {
     expect(screen.queryByRole('button', { name: '编辑' })).toBeNull()
     fireEvent.click(screen.getByRole('button', { name: '复制' }))
     expect(writeText).toHaveBeenCalledWith('hello bubble')
+    fireEvent.click(screen.getByRole('button', { name: '在新对话中分支' }))
+    expect(onFork).toHaveBeenCalledWith(1)
   })
 
   it('user copy falls back to execCommand when clipboard.writeText is unavailable', () => {
@@ -410,12 +414,15 @@ describe('small branch tails', () => {
     })
     const now = new Date()
     const time = new Date(now.getFullYear(), now.getMonth(), now.getDate(), 14, 24).getTime()
+    const onFork = vi.fn()
     const settled = render(
       <AssistantMarkdown
         t={t}
         blocks={[{ kind: 'text', text: 'answer body' }, { kind: 'reasoning', text: 'hidden' }]}
         streaming={false}
         time={time}
+        seq={3}
+        onFork={onFork}
       />,
     )
     expect(settled.getByText('14:24')).toBeTruthy()
@@ -423,6 +430,8 @@ describe('small branch tails', () => {
     expect(settled.getByRole('button', { name: '在新对话中分支' })).toBeTruthy()
     fireEvent.click(settled.getByRole('button', { name: '复制' }))
     expect(writeText).toHaveBeenCalledWith('answer body')
+    fireEvent.click(settled.getByRole('button', { name: '在新对话中分支' }))
+    expect(onFork).toHaveBeenCalledWith(3)
     settled.unmount()
 
     const thinkOnly = render(

packages/client/ui-conversation/tests/chat-code-subcalls.spec.tsx

@@ -67,7 +67,7 @@ function snapshotWith(
   runningCalls: RunningToolCall[] = [],
 ): ConversationSnapshot {
   return {
-    sessionId: SID, nodes, partial: null, runningCalls, codeDispatches,
+    sessionId: SID, nodes, turnEnds: new Map(), partial: null, runningCalls, codeDispatches,
     pending: [], queue: [], running: runningCalls.length > 0, composerPhase: 'active', removed: false,
     openState: 'open', openError: null,
     hasMore: false, loadingOlder: false, promptError: null, blank: false, subagent: null, lastAgentError: null,

packages/client/ui-conversation/tests/chat-stats-bash-sample.spec.tsx

@@ -32,7 +32,7 @@ const assistant = (seq: number, turn: number, usage?: unknown): AssistantMessage
 
 function snapshotBase(): ConversationSnapshot {
   return {
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: false, composerPhase: 'active', removed: false, openState: 'open', openError: null,
     hasMore: false, loadingOlder: false, promptError: null, blank: false, subagent: null, lastAgentError: null,
   }

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

@@ -20,7 +20,7 @@ import { zh as commonZh } from '@deepseek-ai/dsh-client-locale/src/locales/zh.ts
 import { createChatStore } from '../src/client/stores.ts'
 import { ChatView } from '../src/client/chat/ChatView.tsx'
 import { zh } from '../src/client/locales.ts'
-import { assistantActionsSeqs, deriveChatFlow, flowKeys } from '../src/client/chat/chat-flow.ts'
+import { assistantActionsSeqs, deriveChatFlow, flowKeys, messageBranchSeqs } from '../src/client/chat/chat-flow.ts'
 
 afterEach(cleanup)
 // Keyless create() persists under the bare declared key; clear between cases
@@ -33,7 +33,7 @@ const SID = 's1' as SessionId
 
 function snapshotBase(): ConversationSnapshot {
   return {
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: false, composerPhase: 'active', removed: false, openState: 'open', openError: null,
     hasMore: false, loadingOlder: false, promptError: null, blank: false, subagent: null, lastAgentError: null,
   }
@@ -211,6 +211,24 @@ describe('chat-flow derivation', () => {
     ])
     expect([...seqs].sort((a, b) => a - b)).toEqual([5, 7])
   })
+
+  it('messageBranchSeqs keeps only message rows at completed transcript tails', () => {
+    const interruptedThink: AssistantMessageNode = {
+      kind: 'assistant', seq: 4.1, time: 4_100, turn: 1, step: 2,
+      blocks: [{ kind: 'reasoning', text: 'bad path' }], interrupted: true,
+    }
+    const nodes = [
+      user(1, 'first'),
+      assistant(2, 'answer before tools'),
+      toolResult(3, 'a'),
+      interruptedThink,
+      user(6, 'second'),
+      assistant(7, 'clean tail', 2),
+      user(10, 'user-only tail'),
+    ]
+    const seqs = messageBranchSeqs(nodes, new Map([[1, 5], [2, 8], [3, 11]]))
+    expect([...seqs]).toEqual([7, 10])
+  })
 })
 
 describe('ChatView', () => {
@@ -323,21 +341,38 @@ describe('ChatView', () => {
         user(5, 'next'),
         assistant(6, 'second turn', 2),
       ],
+      turnEnds: new Map([[1, 4], [2, 6]]),
     })
     const view = render(<h.ChatView {...h.props} />)
-    // 2 user + 2 turn-tail assistants; mid-turn text at seq 2 stays chrome-free.
+    // User rows keep copy/clock, while only the two completed assistant tails may branch.
     expect(view.getAllByRole('button', { name: '复制' })).toHaveLength(4)
-    expect(view.getAllByRole('button', { name: '在新对话中分支' })).toHaveLength(4)
+    expect(view.getAllByRole('button', { name: '在新对话中分支' })).toHaveLength(2)
   })
 
-  it('forks from both user and finalized assistant message actions at their event seq', () => {
-    const h = makeHarness({ nodes: [user(1, 'question'), assistant(2, 'answer')] })
+  it('forks only from a finalized assistant at the completed transcript tail', () => {
+    const h = makeHarness({
+      nodes: [user(1, 'question'), assistant(2, 'answer')],
+      turnEnds: new Map([[1, 3]]),
+    })
     const view = render(<h.ChatView {...h.props} />)
     const buttons = view.getAllByRole('button', { name: '在新对话中分支' })
-    expect(buttons).toHaveLength(2)
+    expect(buttons).toHaveLength(1)
     fireEvent.click(buttons[0]!)
-    fireEvent.click(buttons[1]!)
-    expect(h.forkAt.mock.calls).toEqual([[1], [2]])
+    expect(h.forkAt.mock.calls).toEqual([[2]])
+  })
+
+  it('keeps copy chrome but hides branch when tool and interrupted Think follow the response', () => {
+    const interruptedThink: AssistantMessageNode = {
+      kind: 'assistant', seq: 4.1, time: 4_100, turn: 1, step: 2,
+      blocks: [{ kind: 'reasoning', text: 'bad path' }], interrupted: true,
+    }
+    const h = makeHarness({
+      nodes: [user(1, 'question'), assistant(2, 'answer'), toolResult(3, 'a'), interruptedThink],
+      turnEnds: new Map([[1, 5]]),
+    })
+    const view = render(<h.ChatView {...h.props} />)
+    expect(view.getAllByRole('button', { name: '复制' })).toHaveLength(2)
+    expect(view.queryByRole('button', { name: '在新对话中分支' })).toBeNull()
   })
 
   it('renders assistant Markdown across history, streaming, final, and interrupted states while user text stays literal', () => {

packages/client/ui-conversation/tests/diff-card.spec.tsx

@@ -335,7 +335,7 @@ describe('DetailsPanel diff Output section', () => {
 
   function snapshot(over: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
     return {
-      sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+      sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
       pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
       openState: 'open', openError: null, hasMore: false, loadingOlder: false,
       promptError: null, blank: false, subagent: null, lastAgentError: null, ...over,

packages/client/ui-conversation/tests/gate-branch-tails.spec.tsx

@@ -24,7 +24,7 @@ const SID = 's1' as SessionId
 
 function snapshotBase(): ConversationSnapshot {
   return {
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: false, composerPhase: 'active', removed: false, openState: 'open', openError: null,
     hasMore: false, loadingOlder: false, promptError: null, blank: false, subagent: null, lastAgentError: null,
   }

packages/client/ui-conversation/tests/input-bar.spec.tsx

@@ -23,7 +23,7 @@ const SID = 's1' as SessionId
 
 function snapshotOf(overrides: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
   return {
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
     openState: 'open', openError: null, hasMore: false, loadingOlder: false,
     promptError: null, blank: false, subagent: null, lastAgentError: null,

packages/client/ui-conversation/tests/input-matrix.spec.tsx

@@ -26,7 +26,7 @@ const SID = 's1' as SessionId
 /** Standard-props InputBar mount over a real shell (the composer-bar entry shape). */
 function mountBar(shell: SessionInputShell, over?: { running?: boolean; disabled?: boolean }) {
   const session = createSnapshotStore<ConversationSnapshot>({
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: over?.running ?? false, composerPhase: 'active',
     removed: over?.disabled ?? false, openState: 'open', openError: null, hasMore: false,
     loadingOlder: false, promptError: null, blank: false, subagent: null, lastAgentError: null,

packages/client/ui-conversation/tests/input-scenarios.spec.tsx

@@ -112,7 +112,7 @@ async function scopedBench(register?: (slash: SlashService) => void) {
   actx.on('slash/input-consume-token', req => shell.consumeToken(req.guard) ? true : undefined)
   const wiring = shell
   const sessionStore = createSnapshotStore<ConversationSnapshot>({
-    sessionId, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
     openState: 'open', openError: null, hasMore: false, loadingOlder: false,
     promptError: null, blank: false, subagent: null, lastAgentError: null,

packages/client/ui-conversation/tests/queue-dock.spec.tsx

@@ -28,7 +28,7 @@ function row(id: string, text: string | null, preview = text ?? '[image]'): Queu
 
 function snapshotWith(queue: QueuedMessage[]): ConversationSnapshot {
   return {
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue, running: true, composerPhase: 'active', removed: false, openState: 'open', openError: null,
     hasMore: false, loadingOlder: false, promptError: null, blank: false, subagent: null, lastAgentError: null,
   }

packages/client/ui-conversation/tests/read-card.spec.tsx

@@ -283,7 +283,7 @@ describe('DetailsPanel Output section (read)', () => {
 
   function snapshot(over: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
     return {
-      sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+      sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
       pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
       openState: 'open', openError: null, hasMore: false, loadingOlder: false,
       promptError: null, blank: false, subagent: null, lastAgentError: null, ...over,

packages/client/ui-conversation/tests/search-card.spec.tsx

@@ -397,7 +397,7 @@ describe('DetailsPanel Output section (search)', () => {
 
   function snapshot(over: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
     return {
-      sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+      sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
       pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
       openState: 'open', openError: null, hasMore: false, loadingOlder: false,
       promptError: null, blank: false, subagent: null, lastAgentError: null, ...over,

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

@@ -68,7 +68,7 @@ const workspaceState = (items: readonly WorkspaceView[]): WorkspaceListState =>
 
 function conversationSnapshot(overrides: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
   return {
-    sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+    sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
     pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
     openState: 'open', openError: null, hasMore: false, loadingOlder: false,
     promptError: null, blank: false, subagent: null, lastAgentError: null,

packages/client/ui-conversation/tests/terminal-card.spec.tsx

@@ -455,7 +455,7 @@ describe('DetailsPanel Output section', () => {
 
   function snapshot(over: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
     return {
-      sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+      sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
       pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
       openState: 'open', openError: null, hasMore: false, loadingOlder: false,
       promptError: null, blank: false, subagent: null, lastAgentError: null, ...over,

packages/client/ui-conversation/tests/web-card.spec.tsx

@@ -232,7 +232,7 @@ describe('DetailsPanel web Output section', () => {
 
   function snapshot(over: Partial<ConversationSnapshot> = {}): ConversationSnapshot {
     return {
-      sessionId: SID, nodes: [], partial: null, runningCalls: [], codeDispatches: new Map(),
+      sessionId: SID, nodes: [], turnEnds: new Map(), partial: null, runningCalls: [], codeDispatches: new Map(),
       pending: [], queue: [], running: false, composerPhase: 'active', removed: false,
       openState: 'open', openError: null, hasMore: false, loadingOlder: false,
       promptError: null, blank: false, subagent: null, lastAgentError: null, ...over,