fix(gui): tail history page carries the full-log todo projection

The client's todos projection derived only from the paged display window,
so reopening a session whose last todo/write preceded the tail page showed
an empty plan until the user paged back — session-level state cannot be
reconstructed from an arbitrary window. The host owns the full log, so the
tail history response now attaches todos (latest todo/write backscan, the
same posture as the view pairing); installWindow seeds it, window rebuilds
preserve it, and any in-window or live write keeps overwriting it. The
fixture mirrors the host; docs and both Agent Notes record the mechanism.
This commit is contained in:
Chinesezjc
2026-07-24 22:56:12 +08:00
parent 356be9710a
commit 1687c2c15c
12 files changed
+107 -21

No files matched your search

@@ -18,7 +18,7 @@ The model sends the ENTIRE list every call; the new list replaces the old (last-
### State on the session log, not a service
The list is appended as a `todo/write` event carrying the full `{ todos }` snapshot. The harness is event-sourced — the LLM history, tool calls, and turn structure all live on the log — so the todo list lives there too. This buys durability, replay, and `session/load` reconstruction for free: a reopened session re-derives the current list (the last `todo/write`) and the ACP bridge re-emits the `plan` on load, with no separate persistence backend, no in-memory service to rehydrate, and no extra wiring. An in-memory `ctx.todos` service would have had to reinvent all of that. (Full-log consumers get this reconstruction outright; the web client's paged window currently sees only the tail page — the gap and its fix directions are recorded in the [web todo display note](2026-07-23-web-todo-display.md).)
The list is appended as a `todo/write` event carrying the full `{ todos }` snapshot. The harness is event-sourced — the LLM history, tool calls, and turn structure all live on the log — so the todo list lives there too. This buys durability, replay, and `session/load` reconstruction for free: a reopened session re-derives the current list (the last `todo/write`) and the ACP bridge re-emits the `plan` on load, with no separate persistence backend, no in-memory service to rehydrate, and no extra wiring. An in-memory `ctx.todos` service would have had to reinvent all of that. (Full-log consumers get this reconstruction outright; the web client's paged window gets it from the tail history page's host-computed projection — see the [web todo display note](2026-07-23-web-todo-display.md).)
### NOT a surface event
@@ -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
2026-07-23-web-todo-display.md: 003ab29546e8098331496641a28d095a1dd95fed
2026-07-23-web-todo-display.zh.md: 200fd1d44dca7c5bcaa4f47ac31e26cb1f0f8d3d
2026-07-23-web-todo-display.md: a5b45f288518cd53594aea724fb8eba532a1142d
2026-07-23-web-todo-display.zh.md: d960a5ab402d9f1c53ddecc7838b19da0743ef3b
@@ -32,4 +32,4 @@ The dedicated `todo_write` chat row registers through the named `ctx.toolviews`
## Consequences
Replay correctness is owned by one code path: any future change to window rebuild keeps todos consistent for free, and the fixture (fx-alpha turn 63) plus `scripts/verify-todo-display.mjs` pin the full chain (panel visibility, row summary, details linkage, collapse, dark theme) in a real chromium. `todos` is a required `ConversationSnapshot` field, so scripted fakes in specs must carry it. The ACP bridge's todo → `plan` mapping and the TUI panel are untouched; the web surfaces render the same event with no new wire vocabulary. Known gap: the projection is window-scoped — reopening a session whose last `todo/write` precedes the tail history page shows an empty plan until the user pages back to it; restoring the tool note's cold-load reconstruction promise needs the current projection independent of the display window (host-attached on the history response, or a dedicated read).
Replay correctness is owned by one code path: any future change to window rebuild keeps todos consistent for free, and the fixture (fx-alpha turn 63) plus `scripts/verify-todo-display.mjs` pin the full chain (panel visibility, row summary, details linkage, collapse, dark theme) in a real chromium. `todos` is a required `ConversationSnapshot` field, so scripted fakes in specs must carry it. The ACP bridge's todo → `plan` mapping and the TUI panel are untouched; the web surfaces render the same event with no new wire vocabulary. Cold-load reconstruction is host-backed: the tail history page carries `todos` — the full-log latest `todo/write`, computed independently of the page window (the same backscan posture the view pairing uses) — so a reopened session restores the plan even when the last write precedes the window; the seeded value is preserved across window rebuilds and overwritten by any later write.
@@ -32,4 +32,4 @@ Status: implemented
## Consequences
回放正确性由一条代码路径掌管:未来对窗口重建的任何改动都免费保持 todos 一致;fx-alpha 第 63 轮的 fixture(测试前置数据)加 `scripts/verify-todo-display.mjs` 在真实 chromium 里钉住整条链(面板可见性、行摘要、详情联动、折叠、深色主题)。`todos``ConversationSnapshot` 的必填字段,所以 spec 里脚本化的 fake 必须带上它。ACP 桥接的 todo → `plan` 映射与 TUI 面板均未受改动;Web 各面渲染同一个事件,不引入任何新的协议词汇。已知缺口:该投影以窗口为界——重新打开一个最后一次 `todo/write` 落在尾页之前的会话时,计划面板为空,直到用户翻页翻到它;要兑现工具 Note 里冷加载重建的承诺,需要一份独立于显示窗口的当前投影(history 响应由 host 附带,或提供专门的读取口)
回放正确性由一条代码路径掌管:未来对窗口重建的任何改动都免费保持 todos 一致;fx-alpha 第 63 轮的 fixture(测试前置数据)加 `scripts/verify-todo-display.mjs` 在真实 chromium 里钉住整条链(面板可见性、行摘要、详情联动、折叠、深色主题)。`todos``ConversationSnapshot` 的必填字段,所以 spec 里脚本化的 fake 必须带上它。ACP 桥接的 todo → `plan` 映射与 TUI 面板均未受改动;Web 各面渲染同一个事件,不引入任何新的协议词汇。冷加载重建由 host 兜底:history 尾页附带 `todos`——全量 log 上最新一次 `todo/write` 的投影,独立于分页窗口计算(与 view 配对同一种 backscan 姿势)——因此重开会话时即使最后一次写入落在窗口之前,计划也照常恢复;播种值跨窗口重建保留,之后的任何写入照常覆盖
@@ -6,7 +6,7 @@
// approval/question requests exercise replay and composer takeover with stable rpcIds.
import type { ContentBlock } from '@deepseek-ai/dsh-llm/types'
import type { SessionEvent, SessionId } from '@deepseek-ai/dsh-session/types'
import type { SessionEvent, SessionId, TodoItem } from '@deepseek-ai/dsh-session/types'
import type {
ApiProxy, ClientRequest, ClientResponse, HistoryEntry, HostFrame, MuxFrame, RpcReceipt,
RpcRequest, RpcResponse, RpcResult, ServerRequest, ServerResponse, SessionSummary,
@@ -254,6 +254,15 @@ function pageOf(
return { events, hasMore: start > 0 }
}
/** Current todo projection over the full log (host parallel: latest todo/write, last write wins). */
function backscanTodos(log: readonly SessionEvent[]): TodoItem[] | undefined {
for (let i = log.length - 1; i >= 0; i--) {
const event = log[i]
if (event !== undefined && event.type === 'todo/write') return event.data.todos
}
return undefined
}
interface StreamConn<F> {
push(envelope: RpcRequest<F>): void
}
@@ -492,12 +501,14 @@ export function createFixtureApi(): ApiProxy {
const log = logs.get(request.payload.sessionId) ?? []
// Snapshot at request time, deliver after the transit delay (mirrors a real host under latency).
const page = pageOf(log, request.payload.beforeSeq, request.payload.maxMessages ?? 50)
// Tail page carries the session-level todo projection (host parallel: full-log backscan).
const todos = request.payload.beforeSeq === undefined ? backscanTodos(log) : undefined
const doomed = failNextHistory
failNextHistory = false
const delay = historyDelayMs
if (delay > 0) await new Promise(resolve => setTimeout(resolve, delay))
if (doomed) throw new Error('fixture: simulated history transport failure')
return ok(request, page)
return ok(request, { ...page, ...todos === undefined ? {} : { todos } })
},
prompt: (request) => {
const { sessionId: id, mode, content } = request.payload
+2 -2
View File
@@ -1,6 +1,6 @@
# @deepseek-ai/dsh-client-runtime
Client cordis boot + core services: SlotsService (Service wrapper over SlotCore + 'slots/changed' bridge), SessionsService (list store projection, scope tree, bindings, ancestry), the Session object layer (exported as a type; instances are owned and handed out by SessionsService — the manager/paging internals stay package-internal, tests reach them via src), ClientLoader (`./loader` subpath, statically held by the shell). Contract: api-contracts v3 §4. `ConversationSnapshot` carries `todos` — the latest `todo/write` whole-list snapshot in the window, consumed as a per-event side effect (last write wins) and reset on every window rebuild.
Client cordis boot + core services: SlotsService (Service wrapper over SlotCore + 'slots/changed' bridge), SessionsService (list store projection, scope tree, bindings, ancestry), the Session object layer (exported as a type; instances are owned and handed out by SessionsService — the manager/paging internals stay package-internal, tests reach them via src), ClientLoader (`./loader` subpath, statically held by the shell). Contract: api-contracts v3 §4. `ConversationSnapshot` carries `todos` — the session-level todo projection: seeded from the tail history page's full-log projection (`history` response `todos`), overwritten by every in-window or live `todo/write` (last write wins), and preserved across window rebuilds.
## Session title projection
@@ -19,4 +19,4 @@ None; this package neither assembles nor sends a provider request.
- **`loader.unload` is a stub (throws not-implemented)** — the full chain (fiber dispose → registration cascade → style removal) lands with the HMR project.
- **Scope teardown is stage-driven, single-occupant today** — the staged session follows `list.current` exactly (staging is the open signal: the event window opens ⟺ the session is on stage); a removed-while-staged session's scope survives frozen until the stage moves on, not until true observer count reaches zero. Resolution (`cell()`/`binding()`/`scope()`) is pure addressing, render-safe. The staged state can widen to a multi-pane list when concurrent panes land.
- **Value imports of this package from plugin bundles must use the `/client` subpath** — the bare package name is not in the loader externals table and inlines a second module instance, whose private scope-tag Symbol never matches (the empty-state P0 postmortem).
- **`todos` is window-scoped** — the projection scans only the paged display window, so reopening a session whose last `todo/write` precedes the tail page shows an empty plan until the user pages back to it. Restoring the tool's cold-load reconstruction promise needs the current projection independent of the window (host-attached on history, or a dedicated read).
- **`SessionSummary.title` is a display projection** — the wire summary carries no title yet; the cwd basename stands in, then the raw id.
@@ -327,13 +327,13 @@ export class Session implements ObservableSnapshot<ConversationSnapshot> {
this.openError = result.error
return
}
this.installWindow(result.value.events, result.value.hasMore)
this.installWindow(result.value.events, result.value.hasMore, result.value.todos)
// Gap detection (§D.3-4): baseline past the window tail and liveBuffer did not cover it -> pull the tail page once more.
const tailSeq = this.windowTailSeq()
if (this.subscribedLastSeq !== null && tailSeq !== null && this.subscribedLastSeq > tailSeq) {
result = (await this.api.sessions.history({ sessionId: this.sessionId, maxMessages: PAGE_MESSAGES })).result
if (generation !== this.openGeneration) return
if (result.ok) this.installWindow(result.value.events, result.value.hasMore)
if (result.ok) this.installWindow(result.value.events, result.value.hasMore, result.value.todos)
}
this.openState = 'open'
} catch (error) {
@@ -351,11 +351,15 @@ export class Session implements ObservableSnapshot<ConversationSnapshot> {
* Stitching MUST NOT route through acceptLiveEvent: openState is still 'loading' here
* (doOpen flips it after install), so recursing would push every buffered event straight
* back into liveBuffer where nothing ever drains it — a silent drop loop (audit S1). */
private installWindow(entries: HistoryEntry[], hasMore: boolean): void {
private installWindow(entries: HistoryEntry[], hasMore: boolean, todos?: readonly TodoItem[]): void {
this.events = entries.map(e => e.event)
this.views = entries.map(e => e.view)
this.baseSeq = this.events[0]?.seq ?? 0
this.hasMore = hasMore
// Session-level projection from the tail page (full-log latest todo/write,
// independent of the window); an in-window write below re-derives the same
// value, and later live events keep overwriting it.
if (todos !== undefined) this.todos = todos
this.foldAdapter.reset(this.events, this.baseSeq, this.views)
this.rebuildDerivedFromWindow()
const buffered = this.liveBuffer
@@ -494,14 +498,16 @@ export class Session implements ObservableSnapshot<ConversationSnapshot> {
/** Re-derive state (partial/openCalls/frozenNodes) from raw window events after a rebuild — keeps
* paging/stitching consistent, and makes the live freeze and the history replay converge on the
* same interrupted nodes (chunks are logged, so the replayed sweep re-freezes identical text). */
* same interrupted nodes (chunks are logged, so the replayed sweep re-freezes identical text).
* todos is deliberately NOT reset: it is session-level (seeded by the tail page's full-log
* projection, not derivable from an arbitrary window). The window always extends to the log
* tail, so an in-window todo/write can only overwrite it with the same latest value. */
private rebuildDerivedFromWindow(): void {
this.partial = null
this.openCalls.clear()
this.callsRev++
this.frozenNodes = []
this.frozenRev++
this.todos = []
for (let i = 0; i < this.events.length; i++) {
const event = this.events[i]
/* v8 ignore next -- dense-array guard: i stays within events.length, so the undefined arm needs a sparse array no caller builds. */
+1 -1
View File
@@ -47,7 +47,7 @@ export class FakeApiClient implements IApiClient {
onList: (payload: unknown) => Promise<RpcResponse<{ items: never[] }>> = () => Promise.resolve(ok({ items: [] }))
onCreate: (payload: unknown) => Promise<RpcResponse<{ sessionId: SessionId }>> = () => Promise.resolve(ok({ sessionId: 'fk-new' as SessionId }))
onHistory: (payload: { sessionId: SessionId; beforeSeq?: number; maxMessages?: number })
=> Promise<RpcResponse<{ events: never[]; hasMore: boolean }>> =
=> Promise<RpcResponse<{ events: never[]; hasMore: boolean; todos?: { content: string; status: 'pending' | 'in_progress' | 'completed' }[] }>> =
() => Promise.resolve(ok({ events: [], hasMore: false }))
onPrompt: (payload: unknown) => Promise<RpcResponse<{ accepted: true }>> = () => Promise.resolve(ok({ accepted: true as const }))
+21 -2
View File
@@ -22,9 +22,9 @@ function makeSession(api = new FakeApiClient()): { api: FakeApiClient; session:
return { api, session: new Session(SID, api) }
}
function histResponse(events: SessionEvent[], hasMore = false) {
function histResponse(events: SessionEvent[], hasMore = false, todos?: { content: string; status: 'pending' | 'in_progress' | 'completed' }[]) {
// history now returns HistoryEntry[] ({event, view?}); these tests are view-less.
return Promise.resolve(ok({ events: entries(events) as never[], hasMore }))
return Promise.resolve(ok({ events: entries(events) as never[], hasMore, ...todos === undefined ? {} : { todos } }))
}
describe('open', () => {
@@ -170,6 +170,25 @@ describe('live event path', () => {
expect(replayed.session.getSnapshot().todos).toEqual(listB)
})
it('seeds todos from the tail page projection when the last write precedes the window', async () => {
const list = [{ content: '窗口外的计划', status: 'in_progress' as const }]
// Cold open: the page window carries NO todo/write; the projection rides the response.
const { api, session } = makeSession()
api.onHistory = () => histResponse(plainTurn(100, 9, '问', '答'), true, list)
await session.open()
expect(session.getSnapshot().todos).toEqual(list)
// Paging an older window in must not clear the session-level projection.
api.onHistory = () => histResponse(plainTurn(94, 8, '旧问', '旧答'), false)
await session.loadOlder()
expect(session.getSnapshot().todos).toEqual(list)
// A later live write still overrides the seeded projection.
session.handleMuxEnvelope('r' as never, {
type: 'session/event', sessionId: SID,
event: ev.todoWrite(106, [{ content: '新计划', status: 'pending' as const }]),
})
expect(session.getSnapshot().todos).toEqual([{ content: '新计划', status: 'pending' }])
})
it('repairs a seq gap by repulling the tail page instead of appending a hole', async () => {
const { api, session } = await opened(plainTurn(0, 0, 'a', 'b')) // tail seq = 5
const repaired = [...plainTurn(0, 0, 'a', 'b'), ...plainTurn(6, 1, 'c', 'd')]
+6 -2
View File
@@ -5,7 +5,7 @@
*/
import type { ContentBlock } from '@deepseek-ai/dsh-llm/types'
import type { SessionEvent, SessionId } from '@deepseek-ai/dsh-session/types'
import type { SessionEvent, SessionId, TodoItem } from '@deepseek-ai/dsh-session/types'
import type { RpcId, RpcRequest, RpcResponse } from './rpc.ts'
import type { ToolEventView } from './events.ts'
@@ -60,9 +60,13 @@ export interface SessionsApi {
* Each entry pairs the raw SessionEvent with the host-computed view (tool events whose
* presenter produced one, evaluated against the registry at pagination time); the client
* rebuilds the surface from the events with the shared fold.
* The tail page (beforeSeq absent) also carries `todos` — the session's current todo
* projection (latest `todo/write` over the FULL log, independent of the page window) —
* so a paged client restores the plan without walking history; absent when the session
* never wrote one. Older pages omit it (the projection is session-level, not per-page).
*/
history(request: RpcRequest<{ sessionId: SessionId; beforeSeq?: number; maxMessages?: number }>):
Promise<RpcResponse<{ events: HistoryEntry[]; hasMore: boolean }>>
Promise<RpcResponse<{ events: HistoryEntry[]; hasMore: boolean; todos?: TodoItem[] }>>
/** Sends a message. content is core's ContentBlock[] verbatim; mode maps 1:1 — queue→send, steer→steer. */
prompt(request: RpcRequest<{ sessionId: SessionId; mode: 'queue' | 'steer'; content: ContentBlock[] }>):
+15 -2
View File
@@ -8,7 +8,7 @@ import { mkdir, stat } from 'node:fs/promises'
import type { Context } from 'cordis'
import type { Agent, AgentStatus } from '@deepseek-ai/dsh-agent'
import type { ContentBlock, MessageSource } from '@deepseek-ai/dsh-llm'
import type { JsonValue, Session, SessionEvent, SessionHeader, SessionId } from '@deepseek-ai/dsh-session'
import type { JsonValue, Session, SessionEvent, SessionHeader, SessionId, TodoItem } from '@deepseek-ai/dsh-session'
import type { SessionPersistence } from '@deepseek-ai/dsh-session-persistence'
import { foldSessionTitle } from '@deepseek-ai/dsh-session-title'
import type {
@@ -265,6 +265,15 @@ function backscanArgs(events: readonly SessionEvent[], callId: string): { name:
return undefined
}
/** Current todo projection: the latest `todo/write` over the full log (whole-list replace ⇒ last write wins); undefined when none. */
function backscanTodos(events: readonly SessionEvent[]): TodoItem[] | undefined {
for (let i = events.length - 1; i >= 0; i--) {
const event = events[i]
if (event !== undefined && event.type === 'todo/write') return event.data.todos
}
return undefined
}
/**
* Thrown by the cold-resume path when the id names no servable session
* (absent from the store, or a pre-project legacy log without a cwd).
@@ -435,7 +444,11 @@ export function createApiProxy(ctx: Context, defaults: ApiProxyDefaults): ApiPro
const view = viewFor(ctx, event, callId => backscanArgs(page.events, callId))
return { event, ...view === undefined ? {} : { view } }
})
return ok(request, { events: entries, hasMore: page.hasMore })
// Tail page carries the session-level todo projection over the FULL
// log (the page window may not contain the last todo/write; a paged
// client cannot reconstruct session-level state from it).
const todos = beforeSeq === undefined ? backscanTodos(found.agent.session.events) : undefined
return ok(request, { events: entries, hasMore: page.hasMore, ...todos === undefined ? {} : { todos } })
},
async prompt(request) {
@@ -154,6 +154,39 @@ describe('mux live view computation', () => {
expect('view' in (byKey.get('tool/result:h-plain') ?? {})).toBe(false)
})
it('tail page carries the full-log todo projection; older pages and todo-less sessions omit it', async () => {
const { ctx } = await harness()
const api = createApiProxy(ctx, { provider: 'p', model: 'm', cwd: '/tmp' })
const session = ctx.sessions.create()
ctx.agents.register({ id: session.id, session, status: 'idle', ctx } as Agent)
// Superseded write early in the log, latest write later; enough messages to page.
session.append('todo/write', { todos: [{ content: 'old', status: 'pending' }] })
for (let turn = 0; turn < 6; turn++) {
session.append('turn/start', { turn, trigger: { kind: 'message', source: { kind: 'user' } } })
session.append('user/message', { content: [{ type: 'text', text: `q${turn}` }], source: { kind: 'user' } }, { surfaceOp: 'append' })
session.append('assistant/message', { turn, step: 0, content: [{ type: 'text', text: `a${turn}` }], provenance: { provider: 'p', model: 'm' } }, { surfaceOp: 'append' })
session.append('turn/end', { turn, reason: { kind: 'completed' } })
}
session.append('todo/write', { todos: [{ content: 'current', status: 'in_progress' }] })
// Tail page limited to 2 messages: the latest todo/write may or may not sit
// in the window — the projection must come from the FULL log either way.
const tail = await api.sessions.history({ rpcId: RpcId('t-todos'), payload: { sessionId: session.id, maxMessages: 2 } })
if (!tail.result.ok) throw new Error('history failed')
expect(tail.result.value.todos).toEqual([{ content: 'current', status: 'in_progress' }])
// An older page omits the projection (session-level, tail-page-only).
const boundary = tail.result.value.events[0]?.event.seq ?? 0
const older = await api.sessions.history({ rpcId: RpcId('t-todos-2'), payload: { sessionId: session.id, beforeSeq: boundary, maxMessages: 2 } })
if (!older.result.ok) throw new Error('older failed')
expect('todos' in older.result.value).toBe(false)
// A session with no todo/write anywhere omits the field.
const bare = ctx.sessions.create()
ctx.agents.register({ id: bare.id, session: bare, status: 'idle', ctx } as Agent)
const bareTail = await api.sessions.history({ rpcId: RpcId('t-todos-3'), payload: { sessionId: bare.id } })
if (!bareTail.result.ok) throw new Error('bare failed')
expect('todos' in bareTail.result.value).toBe(false)
})
it('drops a disposed session from the live open-call table (result after dispose gets no view)', async () => {
const { ctx } = await harness()
const api = createApiProxy(ctx, { provider: 'p', model: 'm', cwd: '/tmp' })