refactor(agent): name delivery methods by intent

This commit is contained in:
Tianyi Cui
2026-07-24 12:27:20 +08:00
parent 7b7f793ee5
commit 086e454931
39 changed files with 514 additions and 484 deletions

View File

@@ -54,12 +54,12 @@ Turn and step boundaries and the model token stream are durable `session/event`
### Agent interface (`types.ts`)
The handle every plugin programs against:
`Agent` is a structural interface. Public delivery methods name caller intent; the concrete driver keeps queue targeting and wakeup routing private ([decision](../../../.agents/notes/implemented/architecture/2026-07-24-intent-named-agent-delivery.md)). `send()`, `queue()`, and `steer()` return an opaque `AgentMessageId` carried by that FIFO item's `agent/inbox/enqueue`/`dequeue`/`discard` events. Each snapshots content, resolved source, attached contexts, and model-hidden metadata as one detached, deeply frozen lossless-JSON record before notification and enqueue; invalid data throws synchronously. Omitting `options.source` attests direct human input as `{ kind: 'user' }`, so every non-human producer labels its content.
- `agent.send(content, options?)` — the one delivery primitive over the (`target` × `wakeup`) matrix; `Agent` is an abstract class whose `followup`/`steer`/`inject` aliases are fixed-preset delegates to it. It returns the accepted message's opaque `AgentMessageId`, which the message's `agent/inbox/enqueue`/`dequeue`/`discard` events carry so a caller can correlate a queued item with its lifecycle. `target: 'next-turn'` (default) queues one independent FIFO item that, if claimed, becomes the sole ordinary message in its turn; `wakeup` (default `true`) wakes a parked driver, while `wakeup: false` queues without waking. `target: 'next-step'` with `wakeup: true` submits steering, and with `wakeup: false` injects durable context without running the model. Omitting `options.source` attests direct human input as `{ kind: 'user' }` (injection defaults to `{ kind: 'plugin', plugin: '' }`) and may authorize policy consumers, so plugins, schedulers, and other non-human producers provide their own source. Content, resolved source, and `options.contexts` become one detached, deeply frozen lossless-JSON record before `agent/inbox/enqueue` and enqueue; invalid data throws synchronously, and caller or notification-listener in-place mutation cannot change the log or model input. After admission, separate contexts become injected `user/message` events, while prompt-prefix contexts are baked before the effective request in the same `user/message`; a block or replacement of the default additional-context decision can discard them. The [one-send-one-turn Agent Note](../../../.agents/notes/implemented/simplification/2026-07-17-one-send-one-turn.md) owns the turn rationale.
- `agent.followup(content, options?)` — the `next-turn`/wakeup preset of `send()`: queue an ordinary follow-up turn and wake the driver.
- `agent.steer(content, options?)` — the `next-step`/wakeup preset: while running, queue steering for the next checkpoint without dispatching `agent/prompt-submit`; when idle, delegate to a woken follow-up. Attached contexts remain in the same frozen record; separate contexts append immediately after the steering event, while prompt-prefix contexts are baked into that steering event. Both survive late-steering conversion to queued input and disappear with their message on cancellation or terminal discard. Policy can still stop before another step; after turn close and its checkpoint, remaining steering becomes later queued input unless terminal turn policy, cancellation, or disposal discards it.
- `agent.inject(content, options?)` — the `next-step`/no-wakeup preset: accept detached in-session context without running the model; the next request sees its `user/message` (default plugin source) with `content` rendered verbatim as a user-role message. `options.meta` persists opaque JSON state without rendering it. While a turn is open it joins that turn, deferring FIFO while the current tool batch executes and draining before turn close if execution is interrupted; while idle it is wrapped in a one-shot `injection` turn and durability checkpoint ([the turn-enclosure invariant](../../../.agents/notes/implemented/architecture/2026-06-15-turn-enclosure-invariant.md)). Injection bypasses the FIFOs and emits no `agent/inbox/*` event.
- `agent.send(content, options?)` — queue one independent FIFO message as its own turn and wake the driver. After admission, separate contexts become injected `user/message` events, while prompt-prefix contexts are baked before the effective request in the same `user/message`; a block or replacement of the default additional-context decision can discard them. The [one-send-one-turn Agent Note](../../../.agents/notes/implemented/simplification/2026-07-17-one-send-one-turn.md) owns the turn rationale.
- `agent.queue(content, options?)` — queue the same ordinary message without waking an idle driver. A lone queued item leaves `whenIdle()` resolved and rides along before the next waking message.
- `agent.steer(content, options?)` — while running, queue steering for the next checkpoint without dispatching `agent/prompt-submit`; while idle, create a waking ordinary turn. Attached contexts remain in the same frozen record; separate contexts append immediately after the steering event, while prompt-prefix contexts are baked into that event. Both survive late-steering conversion to queued input and disappear with their message on cancellation or terminal discard. Policy can still stop before another step; after turn close and its checkpoint, remaining steering becomes later queued input unless terminal turn policy, cancellation, or disposal discards it.
- `agent.inject(content, options?)` — accept detached in-session context without running the model; the next request sees its `user/message` (default plugin source) with `content` rendered verbatim as a user-role message. `InjectOptions` deliberately has no attached contexts. `options.meta` persists opaque JSON state without rendering it. While a turn is open the injection joins that turn, deferring FIFO while the current tool batch executes and draining before turn close if execution is interrupted; while idle it is wrapped in a one-shot `injection` turn and durability checkpoint ([the turn-enclosure invariant](../../../.agents/notes/implemented/architecture/2026-06-15-turn-enclosure-invariant.md)). Injection bypasses the FIFOs and emits no `agent/inbox/*` event.
- `agent.cancel(cause?, options?)` — cancel the active turn and, unless `options.keepInbox`, ALL pending work: an omitted cause means `{ kind: 'user' }`; TypeScript restricts callers to the `user | parent` union, and an active holder copies its discriminant into a detached frozen signal reason before aborting. An effective call emits `agent/cancel-requested` with the cause before clearing queued and steering work; dropped items are reported on `agent/inbox/discard`, and observers may synchronize state but cannot veto cancellation. `keepInbox: true` aborts the turn but preserves queued and steering items (no discard, and un-started work is not dropped). The same-process typed seam adds no runtime validation or compatibility fallback for untyped callers. Repeated active-turn cancellation is first-wins for the signal, and idle cancellation is a safe no-op with no notification. ACP maps to `user`, while in-process parent propagation maps to `parent`. The cause is runtime-only; durable `turn/end` stays coarse `aborted`.
- `agent.whenIdle()` — resolve once the agent reaches quiescence after settling out of `running` (idle → immediately; disposed → awaits the loop exit). A non-owner's quiescence-observation hook: it observes the work settling WITHOUT tearing the agent down. Teardown is separate — a lifecycle owner stops and unregisters via `AgentHandle.dispose()`, which awaits the loop exit directly.
- `agent.session`, `agent.status`, `agent.options`, `agent.id`

View File

@@ -27,33 +27,11 @@ export interface AgentOptions {
}
/**
* Which inbox queue a {@link Agent.send} item joins:
* - `next-turn` — the item becomes its own turn, claimed at a turn boundary.
* - `next-step` — the item joins the active turn between steps as steering,
* or, when no turn is active, is promoted per its `wakeup` flag.
*/
export type SendTarget = 'next-turn' | 'next-step'
/**
* Options for the unified {@link Agent.send} primitive over the
* (`target` × `wakeup`) matrix. Named presets: {@link Agent.followup}
* (`next-turn`/wakeup), {@link Agent.steer} (`next-step`/wakeup), and
* {@link Agent.inject} (`next-step`/no-wakeup).
*
* Options for {@link Agent.send}, {@link Agent.queue}, and {@link Agent.steer}.
* An omitted source attests direct human input as `{ kind: 'user' }` and may
* authorize policy consumers, so non-human producers must label their content.
*/
export interface SendOptions {
/** Queue the item joins; defaults to `next-turn`. */
target?: SendTarget
/**
* Whether this item makes the model run: wake a parked driver (`next-turn`)
* or force a continuation step (`next-step` while running). Defaults to
* `true`. A `false` `next-turn` item queues without waking; a `false`
* `next-step` item attaches durable context without forcing another step
* (the injection preset).
*/
wakeup?: boolean
source?: MessageSource
/**
* Model-facing contexts captured with this inbox item. A queued prompt exposes
@@ -65,12 +43,17 @@ export interface SendOptions {
meta?: JsonValue
}
/** Options accepted by the fixed-preset aliases, which own `target` and `wakeup`. */
export type AliasSendOptions = Omit<SendOptions, 'target' | 'wakeup'>
/** Options specific to durable synthetic context injection. */
export interface InjectOptions {
/** Defaults to `{ kind: 'plugin', plugin: '' }`; non-human producers should identify themselves. */
source?: MessageSource
/** Opaque JSON state retained on the durable message but hidden from the model. */
meta?: JsonValue
}
/**
* Opaque id assigned to one accepted {@link Agent.send} message; returned by
* `send` and carried on its `agent/inbox/*` events for correlation.
* Opaque id assigned to one accepted agent input. FIFO inputs carry the same id
* on their `agent/inbox/*` events; injection bypasses those events.
*/
export type AgentMessageId = Branded<'AgentMessageId'>
@@ -84,24 +67,24 @@ export function AgentMessageId(id: string): AgentMessageId {
}
/**
* One accepted {@link Agent.send} message, carried by the `agent/inbox/*` live
* events. `id` is the value `send` returned to the caller, stable across this
* message's enqueue, dequeue, and discard events. Source defaults are already
* applied, so these are the exact values the item was accepted with. `steering`
* is true for a `next-step` item drained between steps; a `next-turn` item is
* claimed at a turn boundary. `SendOptions.meta` is intentionally omitted: it is
* durable model-hidden state that lands on the eventual `user/message`/
* One accepted FIFO message, carried by the `agent/inbox/*` live events. `id`
* is the value `send`, `queue`, or `steer` returned to the caller, stable across
* this message's enqueue, dequeue, and discard events. Source defaults are
* already applied, so these are the exact values the item was accepted with.
* `steering` is true for an item drained between steps; otherwise it is claimed
* at a turn boundary. `SendOptions.meta` is intentionally omitted: it is durable
* model-hidden state that lands on the eventual `user/message`/
* `steering/message`, not live-event routing data.
*/
export interface AgentMessage {
/** The id `send` returned for this message. */
/** The id returned by the accepting `send`, `queue`, or `steer` call. */
id: AgentMessageId
content: ContentBlock[]
source: MessageSource
contexts: HookContext[]
/** Whether the item joined the steering FIFO (`next-step`) rather than the queued FIFO. */
/** Whether the item joined the steering FIFO rather than the queued FIFO. */
steering: boolean
/** Whether the item is marked to wake the driver or force a continuation. */
/** Whether the item wakes the driver or requests another step. */
wakeup: boolean
}
@@ -119,7 +102,7 @@ export interface CancelOptions {
* An agent's lifecycle state, emitted on every transition as `agent/status`:
* `idle` (parked, waiting for queued work), `running` (the driver is draining
* work and may be closing or checkpointing a turn), `disposed` (terminal — no
* transition leaves it, and `send`/`followup`/`steer`/`inject` throw).
* transition leaves it, and `send`/`queue`/`steer`/`inject` throw).
*/
export type AgentStatus = 'idle' | 'running' | 'disposed'
@@ -179,46 +162,66 @@ export type AgentCancelCause =
/** Runtime reason carried by the signal that controls one live turn. */
export type AgentInterruptReason = AgentCancelCause | { readonly kind: 'disposed' }
/**
* Public agent handle; its concrete implementation is internal to
* `@deepseek-ai/dsh-agent-loop`. An abstract class rather than an interface so
* the fixed-preset aliases ({@link Agent.followup}, {@link Agent.steer},
* {@link Agent.inject}) are shared concrete delegates over the single abstract
* {@link Agent.send} primitive; concrete drivers implement `send` once.
*/
export abstract class Agent {
/** Public agent handle; its concrete implementation is internal to `@deepseek-ai/dsh-agent-loop`. */
export interface Agent {
/** The single identity shared with {@link session}. */
abstract readonly id: SessionId
readonly id: SessionId
/** The provider route and model this agent's requests use. */
abstract readonly options: AgentOptions
readonly options: AgentOptions
/** The live session this agent drives; its log is the durable source of truth. */
abstract readonly session: Session
readonly session: Session
/** The current lifecycle state, mirrored on every `agent/status` transition. */
abstract readonly status: AgentStatus
readonly status: AgentStatus
/** Agent-scoped context; its contributions are agent-local, unwind on disposal, and reject registration afterward. */
abstract readonly ctx: Context
readonly ctx: Context
/**
* The unified delivery primitive over the (`target` × `wakeup`) matrix.
* Detaches, validates, and freezes one lossless-JSON item, then routes it:
*
* - `next-turn` (default) queues an item that becomes the sole ordinary
* message of its own FIFO-ordered turn; `wakeup` (default `true`) wakes a
* parked driver, while `wakeup:false` queues without waking.
* - `next-step` with `wakeup:true` submits steering into the active turn
* (idle falls back to a woken `next-turn`).
* - `next-step` with `wakeup:false` injects durable model-facing context
* without running the model: an open turn joins at the current log position
* (deferred behind an executing tool batch until it settles), and an idle
* inject records a one-shot turn with its own durability checkpoint.
*
* Attached contexts share the same snapshot and ownership boundary. Invalid
* input throws synchronously before any notification, enqueue, or append.
* @param content - the model-facing content blocks to deliver.
* @param options - target queue, wakeup decision, source, contexts, and meta.
* Queue an ordinary message as its own FIFO-ordered turn and wake the driver.
* Content, resolved source, and attached contexts are detached, validated,
* and frozen together; invalid input throws synchronously before notification
* or enqueue.
* @param content - the prompt content blocks.
* @param options - source, attached contexts, and durable model-hidden meta.
* @returns the accepted message's {@link AgentMessageId}, stable across its `agent/inbox/*` events.
*/
abstract send(content: ContentBlock[], options?: SendOptions): AgentMessageId
send(content: ContentBlock[], options?: SendOptions): AgentMessageId
/**
* Queue an ordinary message without waking an idle driver. The item retains
* FIFO order and is claimed only after another input wakes the driver. A lone
* queued item leaves `whenIdle()` resolved.
* @param content - the prompt content blocks.
* @param options - source, attached contexts, and durable model-hidden meta.
* @returns the accepted message's {@link AgentMessageId}, stable across its `agent/inbox/*` events.
*/
queue(content: ContentBlock[], options?: SendOptions): AgentMessageId
/**
* Submit steering into the running turn and request another step. An open turn
* records it at the next steering checkpoint before a request or continuation
* decision; policy may stop before another step. After turn close and its
* checkpoint, any remainder is queued for a later turn; terminal
* `agent/turn-stop`, cancellation, or disposal may discard it. Idle steering
* becomes a waking ordinary turn.
* @param content - the steering content blocks.
* @param options - source, attached contexts, and durable model-hidden meta.
* @returns the accepted message's {@link AgentMessageId}, stable across its `agent/inbox/*` events.
*/
steer(content: ContentBlock[], options?: SendOptions): AgentMessageId
/**
* Append detached model-facing context without running the model. An open-turn
* injection joins at the current log position unless the current tool batch is
* executing; then it waits FIFO until that batch settles and drains before
* turn close even when interrupted. Idle injection uses a one-shot turn and
* durability checkpoint. Disposal awaits idle checkpoints; flush failures
* report through `agent/error`. An omitted source defaults to
* `{ kind: 'plugin', plugin: '' }`.
* @param content - the injected context content blocks.
* @param options - source and durable model-hidden meta.
* @returns the accepted injection's {@link AgentMessageId}; injection emits no `agent/inbox/*` events.
*/
inject(content: ContentBlock[], options?: InjectOptions): AgentMessageId
/**
* Clear queued and steering work — unless `keepInbox` — and abort the active
@@ -230,53 +233,10 @@ export abstract class Agent {
* @param cause - the stable caller intent carried by the current turn signal.
* @param options - cancellation options; `keepInbox` preserves pending work.
*/
abstract cancel(cause?: AgentCancelCause, options?: CancelOptions): void
cancel(cause?: AgentCancelCause, options?: CancelOptions): void
/** Resolve at idle quiescence; disposal waits for driver exit rather than only the status transition. */
abstract whenIdle(): Promise<void>
/**
* Queue an ordinary follow-up turn and wake the driver — the
* `next-turn`/wakeup preset of {@link send}. The item becomes the sole
* ordinary message of its own turn.
* @param content - the prompt content blocks.
* @param options - source and attached contexts.
* @returns the accepted message's {@link AgentMessageId}.
*/
followup(content: ContentBlock[], options?: AliasSendOptions): AgentMessageId {
return this.send(content, { ...options, target: 'next-turn', wakeup: true })
}
/**
* Submit steering into the running turn — the `next-step`/wakeup preset of
* {@link send}. An open turn records it at the next steering checkpoint before
* a request or continuation decision; policy may stop before another step.
* After turn close and its checkpoint, any remainder is queued for a later
* turn; terminal `agent/turn-stop`, cancellation, or disposal may discard it.
* Idle steering falls back to a woken follow-up turn.
* @param content - the steering content blocks.
* @param options - source and attached contexts.
* @returns the accepted message's {@link AgentMessageId}.
*/
steer(content: ContentBlock[], options?: AliasSendOptions): AgentMessageId {
return this.send(content, { ...options, target: 'next-step', wakeup: true })
}
/**
* Append detached model-facing context without running the model — the
* `next-step`/no-wakeup preset of {@link send}. An open-turn injection joins
* at the current log position unless the current tool batch is executing;
* then it waits FIFO until that batch settles and drains before turn close
* even when interrupted. Idle injection uses a one-shot turn and durability
* checkpoint. Disposal awaits idle checkpoints; flush failures report through
* `agent/error`. An omitted source defaults to `{ kind: 'plugin', plugin: '' }`.
* @param content - the injected context content blocks.
* @param options - source and durable model-hidden meta.
* @returns the accepted message's {@link AgentMessageId}.
*/
inject(content: ContentBlock[], options?: AliasSendOptions): AgentMessageId {
return this.send(content, { ...options, target: 'next-step', wakeup: false })
}
whenIdle(): Promise<void>
}
declare module 'cordis' {
@@ -303,8 +263,8 @@ declare module 'cordis' {
*/
'agent/disposed'(this: Scoped<Agent>, agent: Agent): void
/**
* Agent status changed (`idle` ⇄ `running`, or → `disposed`). `send()` does
* not enter `running` synchronously; drive lifecycle from this event.
* Agent status changed (`idle` ⇄ `running`, or → `disposed`). A waking
* delivery does not enter `running` synchronously; drive lifecycle from this event.
* @param agent - the agent whose status flipped.
* @param status - the status just entered (the transition's destination).
* Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): agent-scoped listeners receive only that agent.
@@ -316,7 +276,7 @@ declare module 'cordis' {
* FIFO). Source defaults are already applied, so `message` holds the exact
* accepted values. This is the enqueue-time live signal; the durable record
* is the eventual `user/message`/`steering/message`. Injection
* (`next-step`/no-wakeup) bypasses the FIFOs and does not emit this.
* through `agent.inject()` bypasses the FIFOs and does not emit this.
* @param agent - the agent whose inbox received the item.
* @param message - the accepted message (its returned `id`, content, source, contexts, steering, and wakeup facts).
* Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): agent-scoped listeners receive only that agent.

View File

@@ -3,32 +3,47 @@ import { Context, Service, symbols } from 'cordis'
import type { Events } from 'cordis'
import { Session, SessionId } from '@deepseek-ai/dsh-session'
import AgentRegistry, {
Agent,
AgentMessageId,
agentEvents,
agentInterruptReasonOf,
} from '@deepseek-ai/dsh-agent'
import type { AgentCancelCause, AgentFactory, ContinuationStop, CreateAgentOptions, ResumeAgentOptions } from '@deepseek-ai/dsh-agent'
import type {
Agent,
AgentCancelCause,
AgentFactory,
ContinuationStop,
CreateAgentOptions,
InjectOptions,
ResumeAgentOptions,
SendOptions,
} from '@deepseek-ai/dsh-agent'
function stubAgent(rawId: string, overrides: Partial<Agent> = {}): Agent {
const id = SessionId(rawId)
// Agent is an abstract class, so its alias methods live on the prototype and
// object spread would drop them; build the full literal and merge overrides.
return Object.assign(Object.create(Agent.prototype) as Agent, {
return {
id,
options: {},
session: new Session(id),
status: 'idle',
ctx: new Context(),
send: () => AgentMessageId('stub'),
queue: () => AgentMessageId('stub'),
steer: () => AgentMessageId('stub'),
inject: () => AgentMessageId('stub'),
cancel() {},
whenIdle() { return Promise.resolve() },
...overrides,
})
}
}
describe('AgentRegistry', () => {
it('keeps concrete delivery routing out of public options', () => {
expectTypeOf<'target' extends keyof SendOptions ? true : false>().toEqualTypeOf<false>()
expectTypeOf<'wakeup' extends keyof SendOptions ? true : false>().toEqualTypeOf<false>()
expectTypeOf<'contexts' extends keyof InjectOptions ? true : false>().toEqualTypeOf<false>()
})
it('allows terminal stop policy to cooperate asynchronously with turn cancellation', () => {
type TurnStopListener = Events['agent/turn-stop']
type AsyncTurnStopListener = () => Promise<ContinuationStop | undefined>