fix(session-title): preserve agent turn outcomes

This commit is contained in:
Tianyi Cui
2026-07-21 15:49:37 +08:00
parent c622a2881d
commit 5d411f9c4e
15 changed files with 176 additions and 22 deletions

View File

@@ -11,6 +11,7 @@ Creates and holds event-sourced `Session` instances. Persistence is intentionall
- `ctx.sessions.create(id?, { seed?, meta? }?)` validates and detaches durable seed/header data, fills the version and id, defaults `createdAt` to now, publishes the session, and binds it to the calling fiber. Persisted reconstruction supplies its original `createdAt`, `seedLength`, and `delegationDepth`.
- `ctx.sessions.flush(session)` dispatches the awaited parallel durability checkpoint through the session's captured scope. Every listener starts and the call waits for all to settle before reporting failure; unpublished, detached, and stale objects reject.
- `ctx.sessions.appendOutOfBand(session, type, data, trigger)` accepts only plugin event types opted into `OutOfBandSessionEventMap`. It appends directly inside an open turn; otherwise it atomically opens a zero-step plugin turn, appends, closes, and flushes. A target failure still closes and flushes the synthetic turn, and detach is deferred until the sequence settles.
- `findLastMessageTurnEnd(events)` pairs message-triggered starts with their ends and returns the latest matched `turn/end`. Outcome consumers use this fold instead of the raw latest turn boundary because a later injection or plugin-owned zero-step turn has its own outcome.
- `ctx.sessions.fork(source, boundary?, childSessionId?): Session` — Resolve a live session object or id, select a seed through the inclusive `boundary` event seq (default: current last event), require that boundary to be `turn/end`, and create a live child session with lineage metadata.
- `ctx.sessions.get(id: SessionId): Session | undefined`
- `ctx.sessions.list(): Session[]`

View File

@@ -27,6 +27,27 @@ export type { SessionSurface, SurfaceFoldReplacement, SurfaceFoldResult } from '
export { foldSurface, isSurfaceEvent, isSurfaceEligibleType } from './surface.ts'
export { canonicalHeader, foldRequestHeader, headerEquals } from './request-header.ts'
/**
* Find the latest closed message-triggered turn, excluding injection and
* plugin-owned zero-step turns.
* @param events - session events, or an owned suffix, to inspect.
* @returns the latest matching turn end, or `undefined`.
*/
export function findLastMessageTurnEnd(
events: readonly SessionEvent[],
): SessionEvent<'turn/end'> | undefined {
const messageTurns = new Set<number>()
let latest: SessionEvent<'turn/end'> | undefined
for (const event of events) {
if (event.type === 'turn/start') {
if (event.data.trigger.kind === 'message') messageTurns.add(event.data.turn)
continue
}
if (event.type === 'turn/end' && messageTurns.delete(event.data.turn)) latest = event
}
return latest
}
declare module 'cordis' {
interface Context {
sessions: SessionStore

View File

@@ -1,7 +1,13 @@
import { describe, expect, expectTypeOf, it, vi } from 'vitest'
import { Context } from 'cordis'
import { CallId } from '@deepseek-ai/dsh-llm'
import SessionStore, { SESSION_FORMAT_VERSION, Session, SessionEvent, SessionId } from '@deepseek-ai/dsh-session'
import SessionStore, {
findLastMessageTurnEnd,
SESSION_FORMAT_VERSION,
Session,
SessionEvent,
SessionId,
} from '@deepseek-ai/dsh-session'
import type { CreateSessionOptions, SessionEventType, SessionHeader, SessionSurface, TodoItem } from '@deepseek-ai/dsh-session'
describe('Session', () => {
@@ -48,6 +54,42 @@ describe('Session', () => {
expect(structuredClone(turnEnd.data.reason)).toEqual({ kind: 'max-tokens' })
})
it('finds the latest message-turn outcome past later non-message turns', () => {
const session = new Session(SessionId('message-turn-outcome'))
expect(findLastMessageTurnEnd(session.events)).toBeUndefined()
session.append('turn/start', {
turn: 1,
trigger: { kind: 'injection', source: { kind: 'plugin', plugin: 'before' } },
})
session.append('context/message', {
content: [{ type: 'text', text: 'before' }],
source: { kind: 'plugin', plugin: 'before' },
}, { surfaceOp: 'append' })
session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
expect(findLastMessageTurnEnd(session.events)).toBeUndefined()
session.append('turn/start', {
turn: 2,
trigger: { kind: 'message', source: { kind: 'user' } },
})
session.append('user/message', {
content: [{ type: 'text', text: 'bounded prompt' }],
source: { kind: 'user' },
}, { surfaceOp: 'append' })
const messageEnd = session.append('turn/end', { turn: 2, reason: { kind: 'max-tokens' } })
session.append('turn/start', {
turn: 3,
trigger: { kind: 'injection', source: { kind: 'plugin', plugin: 'after' } },
})
session.append('context/message', {
content: [{ type: 'text', text: 'after' }],
source: { kind: 'plugin', plugin: 'after' },
}, { surfaceOp: 'append' })
session.append('turn/end', { turn: 3, reason: { kind: 'completed' } })
expect(findLastMessageTurnEnd(session.events)).toBe(messageEnd)
})
it('renders context and steering messages as plain user content', () => {
const session = new Session(SessionId('s2'))
session.append('context/message', {

View File

@@ -12,7 +12,7 @@ The driver follows this sequence:
2. Call `parent.ctx.agents.create` directly, passing the required request signal into the factory's creation transaction.
3. During that transaction's unpublished setup window, install the requested persona, tool restriction, and structured-output runtime.
4. Publish the child, retain the returned `AgentHandle`, and drive one task with `child.send(prompt)` followed by `child.whenIdle()`.
5. Read the child's own last assistant message and terminal turn reason, excluding any fork seed.
5. Read the child's own last assistant message and latest message-triggered turn reason, excluding any fork seed and later plugin-owned zero-step turns.
The child gets the parent's working-directory/session lineage and inherits the parent model unless `request.agentOptions` overrides it. It gets a fresh flat registration scope: parent ownership does not import parent tool restrictions or establish an authority subset.

View File

@@ -10,7 +10,7 @@
import { randomUUID } from 'node:crypto'
import type { Context } from 'cordis'
import type { Agent, AgentOptions } from '@deepseek-ai/dsh-agent'
import { SessionId, type SessionEvent, type TurnEndReason } from '@deepseek-ai/dsh-session'
import { findLastMessageTurnEnd, SessionId, type SessionEvent, type TurnEndReason } from '@deepseek-ai/dsh-session'
import type { ContentBlock } from '@deepseek-ai/dsh-llm'
import { assertSubagentMaxDepth, delegationDepthOf } from '@deepseek-ai/dsh-subagent'
import type { SubagentResult, SubagentRun, SubagentStartRequest, SubagentStopReason } from '@deepseek-ai/dsh-subagent'
@@ -175,7 +175,7 @@ function readResult(
): SubagentResult {
const own = child.session.events.slice(seedLength)
const lastMessage = own.findLast((event): event is SessionEvent<'assistant/message'> => event.type === 'assistant/message')
const lastEnd = own.findLast((event): event is SessionEvent<'turn/end'> => event.type === 'turn/end')
const lastEnd = findLastMessageTurnEnd(own)
const output: ContentBlock[] = lastMessage?.data.content ?? []
const recorded = toStopReason(lastEnd?.data.reason)
// Disposal can tear the owner down before the loop records its ordinary

View File

@@ -6,7 +6,7 @@ import AgentLoop from '@deepseek-ai/dsh-agent-loop'
import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
import * as Invariants from '@deepseek-ai/dsh-invariants'
import SubagentService from '@deepseek-ai/dsh-subagent'
import { MockAdapter, textResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
import { maxTokensResponse, MockAdapter, textResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
import { startInProcessRun } from '../src/index.ts'
type Script = ConstructorParameters<typeof MockAdapter>[0]
@@ -44,6 +44,36 @@ describe('startInProcessRun', () => {
expect(ctx.agents.get(run.id)).toBeUndefined()
})
it('reports the message-turn outcome when a later non-message turn completes during flush', async () => {
const { ctx, parent } = await setup([maxTokensResponse('partial answer')])
let injected = false
ctx.on('session/flush', (session) => {
if (injected || session.header.parentSession === undefined) return
const lastEnd = session.events.findLast(event => event.type === 'turn/end')
if (lastEnd?.type !== 'turn/end' || lastEnd.data.reason.kind !== 'max-tokens') return
injected = true
const turn = lastEnd.data.turn + 1
session.append('turn/start', {
turn,
trigger: { kind: 'injection', source: { kind: 'plugin', plugin: 'late-metadata' } },
})
session.append('context/message', {
content: [{ type: 'text', text: 'late metadata' }],
source: { kind: 'plugin', plugin: 'late-metadata' },
}, { surfaceOp: 'append' })
session.append('turn/end', { turn, reason: { kind: 'completed' } })
})
const run = await startInProcessRun(request(parent), {})
const result = await run.result
const child = ctx.agents.get(run.id)!
expect(child.session.events.findLast(event => event.type === 'turn/end'))
.toMatchObject({ data: { reason: { kind: 'completed' } } })
expect(result.stopReason).toBe('max-tokens')
await run.dispose()
})
it('seeds a forked child but reads only the child-owned output', async () => {
const { ctx, parent } = await setup([textResponse('parent answer'), textResponse('child answer')])
parent.send([{ type: 'text', text: 'parent question' }])

View File

@@ -20,7 +20,7 @@ The plugin answers `shutdown`, disposes SDK-owned agents and subscriptions to qu
## Wire notes
`initialize.serverInfo.name` is the wire-stable `deepseek-harness-sdk-runtime`. A session accepts one in-flight prompt; overlap fails immediately, other sessions remain independent, and the session is reusable after settlement. Persistence roots and persona come from `cordis.yml`.
`initialize.serverInfo.name` is the wire-stable `deepseek-harness-sdk-runtime`. A session accepts one in-flight prompt; overlap fails immediately, other sessions remain independent, and the session is reusable after settlement. `session.finished` reports that prompt's message-triggered turn outcome; later injection or plugin-owned zero-step turns still stream as `session.event` notifications but cannot replace the prompt status. Persistence roots and persona come from `cordis.yml`.
## Model Experience

View File

@@ -10,7 +10,7 @@ import { resolve } from 'node:path'
import type { ContentBlock } from '@deepseek-ai/dsh-llm'
import type { Agent, AgentHandle } from '@deepseek-ai/dsh-agent'
import { carrierKeyOf, type Scoped } from '@deepseek-ai/dsh-scope'
import { SessionId, type TurnEndReason } from '@deepseek-ai/dsh-session'
import { findLastMessageTurnEnd, SessionId, type TurnEndReason } from '@deepseek-ai/dsh-session'
import type SubagentService from '@deepseek-ai/dsh-subagent'
import type { SubagentRunEndInfo } from '@deepseek-ai/dsh-subagent'
import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
@@ -93,7 +93,9 @@ export class HarnessSdkServer {
this.disposers.push(ctx.on('session/event', (session, event) => {
if (event.type === 'turn/end') {
const rec = this.sessions.get(String(session.id))
if (rec) rec.lastTurnEnd = event.data.reason
if (rec && findLastMessageTurnEnd(session.events)?.seq === event.seq) {
rec.lastTurnEnd = event.data.reason
}
}
this.transport.notify('session.event', { sessionId: String(session.id), event })
}))

View File

@@ -7,7 +7,7 @@ import { afterEach, describe, expect, it, vi } from 'vitest'
import { Context } from 'cordis'
import { type Agent, type AgentHandle } from '@deepseek-ai/dsh-agent'
import { SessionId } from '@deepseek-ai/dsh-session'
import SessionStore, { SessionId } from '@deepseek-ai/dsh-session'
import * as agentCore from '@deepseek-ai/dsh-agent-spine-demo'
import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
@@ -215,6 +215,64 @@ describe('HarnessSdkServer', () => {
expect(otherHandle.dispose).toHaveBeenCalledOnce()
})
it('reports the message-turn outcome when a later non-message turn settles before idle', async () => {
const ctx = new Context()
await ctx.plugin(SessionStore)
const transport = new FakeTransport()
const server = new HarnessSdkServer(ctx, transport) as unknown as {
prompt(params: { sessionId: string; contentBlocks: { type: 'text'; text: string }[] }): Promise<unknown>
sessions: Map<string, { handle: AgentHandle; lastTurnEnd: undefined; activePrompt: boolean }>
shutdown(): Promise<Record<string, never>>
}
const session = ctx.sessions.create(SessionId('message-outcome'))
const agent = {
session,
send(content: { type: 'text'; text: string }[]) {
session.append('turn/start', {
turn: 1,
trigger: { kind: 'message', source: { kind: 'user' } },
})
session.append('user/message', {
content,
source: { kind: 'user' },
}, { surfaceOp: 'append' })
session.append('turn/end', { turn: 1, reason: { kind: 'max-tokens' } })
session.append('turn/start', {
turn: 2,
trigger: { kind: 'injection', source: { kind: 'plugin', plugin: 'late-metadata' } },
})
session.append('context/message', {
content: [{ type: 'text', text: 'late metadata' }],
source: { kind: 'plugin', plugin: 'late-metadata' },
}, { surfaceOp: 'append' })
session.append('turn/end', { turn: 2, reason: { kind: 'completed' } })
},
whenIdle: () => Promise.resolve(),
} as unknown as Agent
server.sessions.set('message-outcome', {
handle: { agent, dispose: () => Promise.resolve() },
lastTurnEnd: undefined,
activePrompt: false,
})
await server.prompt({
sessionId: 'message-outcome',
contentBlocks: [{ type: 'text', text: 'bounded prompt' }],
})
expect(transport.notifications.findLast(notification => notification.method === 'session.finished'))
.toEqual({
method: 'session.finished',
params: {
sessionId: 'message-outcome',
status: 'error',
reason: { kind: 'max-tokens' },
},
})
await server.shutdown()
await ctx.fiber.dispose()
})
it('notifies the host when a child session is created with parent lineage', async () => {
const storageDir = await mkdtemp(join(tmpdir(), 'dsh-jsonrpc-subagent-'))
const ctx = await makeHarness(storageDir)