docs(i18n): refresh core data translations after merge
This commit is contained in:
@@ -11,6 +11,13 @@
|
||||
一个流式响应交错包含多种类型的块(文本、推理(reasoning)、多个工具调用)。`index` 将每个 delta 关联到其所属块;`block-end` 携带完整组装好的 `ContentBlock`,消费方无需自行重新组装 delta。这是一个**封闭的**可辨识联合类型:对 `type` 的 `switch` 以 `assertNever` 结尾,因此新增变体会在每个必须处理它的消费方处触发编译错误。
|
||||
|
||||
```ts type-equiv
|
||||
/**
|
||||
* Raw streaming protocol emitted by adapters.
|
||||
* Block indexes correlate interleaved deltas, and `block-end` carries the
|
||||
* assembled block. Adapters emit usage before the terminal finish and nothing
|
||||
* afterward; tool arguments remain raw JSON strings. Failures either throw or
|
||||
* end with `error`/`aborted`, and consumers must handle both paths.
|
||||
*/
|
||||
type StreamChunk =
|
||||
| { type: 'block-start'; index: number; blockType: ContentBlockType }
|
||||
| { type: 'text-delta'; index: number; text: string }
|
||||
@@ -18,7 +25,32 @@ type StreamChunk =
|
||||
| { type: 'tool-call-delta'; index: number; id: CallId; name?: string; argumentsDelta: string }
|
||||
| { type: 'block-end'; index: number; block: ContentBlock }
|
||||
| { type: 'usage'; usage: TokenUsage }
|
||||
| { type: 'finish'; reason: FinishReason }
|
||||
| {
|
||||
type: 'finish'
|
||||
reason: FinishReason
|
||||
/** Adapter-private lossless-JSON state for replaying a successful response. */
|
||||
replayState?: unknown
|
||||
}
|
||||
```
|
||||
|
||||
## `LlmFailure`
|
||||
|
||||
每个抛出的失败或 final-adapter 带内失败都会规范化为一种可序列化、提供方中立的 payload。`providerRetryAfterMs` 是经校验、由提供方请求的正数延迟,而不是重试决策;`ProviderRequestId` 是用于诊断的不透明品牌字符串。
|
||||
|
||||
```ts type-equiv
|
||||
/** Serializable provider-boundary facts; policy decides whether they are retryable. */
|
||||
interface LlmFailure {
|
||||
/** Human-readable provider or transport failure. */
|
||||
readonly message: string
|
||||
/** Stable provider-neutral machine-routing code. */
|
||||
readonly code: string
|
||||
/** HTTP status observed at the provider boundary, when available. */
|
||||
readonly status?: number
|
||||
/** Provider-requested delay in milliseconds, when valid and available. */
|
||||
readonly providerRetryAfterMs?: number
|
||||
/** Opaque provider-issued request identifier for diagnostics. */
|
||||
readonly requestId?: ProviderRequestId
|
||||
}
|
||||
```
|
||||
|
||||
## 适配器契约
|
||||
@@ -27,28 +59,50 @@ type StreamChunk =
|
||||
|
||||
- **`usage` 在 `finish` 之前,`finish` 之后不再有任何分片。** 将两者都推迟到提供方的流结束标记,这样尾部的 usage-only 分片就不会违反顺序。
|
||||
- **工具调用的 `arguments` 全程保持原始 JSON 字符串。** 部分片段通过 `argumentsDelta` 流式传输;如果提供方返回的是已解析的对象,适配器在 `block-end` 时重新序列化为字符串。
|
||||
- **两条许可的错误路径。** 失败可以从 `stream()` 中 THROW(传输/协议错误),**或者**以 `finish {kind:'error'|'aborted'}` 结束流(提供方带内错误,适用于无法在流中途抛出异常的适配器)。消费方必须同时处理*两种*情况。agent loop(智能体循环)将 finish-error/aborted 转化为轮次错误,绝不会为失败的步骤记录一条正常完成的 assistant 消息。
|
||||
- **两条受支持的错误路径,一种事实形状。** 失败可以从 `stream()` 抛出(传输/协议错误),**或者**以 `finish {kind:'error'|'aborted', failure}` 结束流(无法在流中途抛异常的适配器用它表示提供方带内错误)。`LlmError.failure` 携带同一个 `LlmFailure`。final adapter 边界保留被抛出的确切 `Error` 对象,并将不可变事实关联到该调用;agent loop 关闭失败的 step,再把错误、事实与不可变的先前已重试事实提供给 `agent/request-error`。若未恢复,结构化失败会成为 turn 错误,并且该次尝试不会提交正常 assistant 消息或工具副作用。
|
||||
- **一次适配器调用就是一次提供方尝试。** 适配器禁用库重试。agent 层恢复会打开另一个持久、带编号的 step;直接调用 `ctx.llm.stream()` 的调用方仍然只尝试一次。
|
||||
- **提供方停顿在传输层受到时限约束。** 两个已交付的远程适配器都暴露正数且有限的 `streamIdleTimeoutMs`,默认五分钟。watchdog 只在 iterator `next()` 尚未完成时启动,整个请求使用同一个稳定 signal,把自身到期映射为 `TIMEOUT`,并把更早发生的调用方中止保留为 `ABORTED`。
|
||||
- **上下文溢出只有一个规范 code。** 两个 DeepSeek 适配器都通过 `isContextWindowExceededError()` 对提供方的显式细节分类并暴露 `CONTEXT_WINDOW_EXCEEDED`,无论失败以抛出的 HTTP `LlmError` 还是带内 finish error 到达。消费方按 code 路由,绝不依赖提供方文本。
|
||||
- **每个提供方 HTTP 请求都携带应用归属头。** 适配器发送 `attributionHeaders()`(见下文)作为 `User-Agent` 基线,并通过协议级测试加以证明(mock 服务器断言收到的 header,或对基于库的适配器使用库的 header 钩子)。
|
||||
- **回放状态归适配器所有。** 成功的 `finish` 可以携带重建提供方原生响应所需的无损 JSON 状态。除非 `agent/step-result` listener 改写了内容,否则循环会将其与组装后的 assistant 消息一起存储。后续请求中,仅当历史提供方与目标提供方当前注册到完全相同的适配器实例时,`LlmService` 才会传递该状态。该适配器负责校验状态并拥有所有跨模型或跨提供方转换;其他适配器只会收到提供方中立的内容与 provenance,不会收到私有状态。
|
||||
|
||||
这份契约正是两个适配器作为刻意配对存在的原因:`dsh-llm-deepseek`(手写 fetch/SSE(Server-Sent Events))与 `dsh-llm-pi-ai`(通过 `@earendil-works/pi-ai` 访问同一端点)。两套独立内部实现共享一份契约,正是将协议固定下来的方式:基于库的适配器无法在流中途抛出异常,因此它走通了手写适配器可能不会走到的 finish-chunk 错误路径。
|
||||
该契约由两个有意保持独立的实现锁定:`dsh-llm-deepseek`(手写 fetch/SSE)和 `dsh-llm-pi-ai`(通过 `@earendil-works/pi-ai` 实现的通用多提供方适配器)。基于库的适配器覆盖 finish-chunk 错误路径,而传输边界测试证明每个空闲 watchdog 都会停止其实际请求。
|
||||
|
||||
## `AppIdentity`:应用归属
|
||||
|
||||
每个适配器向提供方发送的静态公开应用标识([`packages/llm/llm/src/attribution.ts`](../../packages/llm/llm/src/attribution.ts))。`attributionHeaders(identity?)` 仅将其映射为标准 `User-Agent` header;本契约有意不支持 OpenRouter 特有的应用归属 header。默认的 `APP_IDENTITY` 从 package manifest(元数据清单)获取版本号;每个字段都是公开的产品事实,不含密钥、路径、会话 id 或用户级标识符,且任何请求级信息都不得影响这些值。设计依据见 [Mandatory `User-Agent` attribution](../rfc/implemented/architecture/2026-06-21-mandatory-app-attribution-headers.md)。
|
||||
每个适配器都会向提供方发送的静态公开应用标识([`packages/llm/llm/src/attribution.ts`](../../packages/llm/llm/src/attribution.ts))。`attributionHeaders(identity?)` 只把它映射到标准 `User-Agent` header;该契约有意不支持 OpenRouter 特有的应用归属 header。默认 `APP_IDENTITY` 从包 manifest 获取版本;每个字段都是公开产品事实——不含 secret、路径、session id 或逐用户标识,且任何逐请求信息都不得影响这些值。设计理由见[强制 `User-Agent` 归属](../../.agents/notes/implemented/architecture/2026-06-21-mandatory-app-attribution-headers.md)。
|
||||
|
||||
```ts type-equiv
|
||||
/**
|
||||
* Static public application identity sent to LLM providers.
|
||||
*
|
||||
* Every field is a public product fact, safe on every request: no secrets,
|
||||
* local paths, session ids, prompt text, or per-user identifiers belong here,
|
||||
* and nothing per-request may influence the values.
|
||||
*/
|
||||
interface AppIdentity {
|
||||
/** `User-Agent` product token (lowercase, hyphenated). */
|
||||
product: string
|
||||
/** Product version; sourced from package metadata, never hand-copied. */
|
||||
version: string
|
||||
/** Public home URL of the app, used as the `User-Agent` comment. */
|
||||
url: string
|
||||
}
|
||||
```
|
||||
|
||||
## `TokenUsage`
|
||||
|
||||
单次调用的 token 用量统计。各计数**互不重叠**:`inputTokens` 仅为未缓存的输入;缓存命中的输入单独报告,计费输入是三者之和。如果提供方将缓存命中合并到单一的 prompt 总量中(如 DeepSeek 的 `prompt_tokens`),适配器需将其减回去。
|
||||
逐调用 token 记账。各计数**互不重叠**:`inputTokens` 只包含未缓存输入;缓存输入单独报告,计费输入是三者之和。若提供方把缓存命中折入单一 prompt 总数(如 DeepSeek 的 `prompt_tokens`),适配器会再将其扣除。`reasoningTokens` 存在时只是信息性细节,已经包含在 `outputTokens` 中;汇总时不得重复相加。
|
||||
|
||||
```ts type-equiv
|
||||
/**
|
||||
* Token accounting for one model call (cache fields are optional).
|
||||
*
|
||||
* Counts are DISJOINT: `inputTokens` is uncached input only; cached input is
|
||||
* reported separately as `cacheReadTokens`/`cacheWriteTokens` (billed input =
|
||||
* sum of the three). Adapters whose providers fold cache hits into a total
|
||||
* prompt count (DeepSeek's `prompt_tokens`) subtract them out.
|
||||
*/
|
||||
interface TokenUsage {
|
||||
inputTokens: number
|
||||
outputTokens: number
|
||||
@@ -60,15 +114,99 @@ interface TokenUsage {
|
||||
|
||||
## `BlockAssembler`
|
||||
|
||||
`BlockAssembler`([`packages/llm/llm/src/assembler.ts`](../../packages/llm/llm/src/assembler.ts))是唯一的共享实现,将 `StreamChunk` 流折叠回 `ContentBlock` 列表与最终的 `Message`。agent loop 记录原始分片(保证回放保真度),同时将相同的分片送入 assembler,因此权威日志保留了 token 级细节,而派生消息可确定性地重建。需要组装结果而不想重新实现折叠逻辑的消费方使用它。
|
||||
`BlockAssembler`([`packages/llm/llm/src/assembler.ts`](../../packages/llm/llm/src/assembler.ts))是唯一的共享实现,负责把 `StreamChunk` 流折叠回 `ContentBlock`、usage、finish reason 与 replay state。循环在记录原始 chunk 的同时,把同一批 chunk 送入 assembler,再将组装后的 assistant 内容连同其 provider/model provenance 一起存储。需要组装结果、又不想重新实现 fold 的消费方使用它。
|
||||
|
||||
```ts public-api
|
||||
/**
|
||||
* Incrementally assembles raw {@link StreamChunk}s into complete
|
||||
* {@link ContentBlock}s and a final assistant {@link Message}.
|
||||
*
|
||||
* The agent loop feeds it while logging raw chunks for replay fidelity, then
|
||||
* reads `blocks()` / `message()` / `usage` / `finish` once the stream ends.
|
||||
*
|
||||
* Tolerant of delta-only protocols (no block-start/end); deltas arriving for
|
||||
* an index already closed by `block-end` are ignored (malformed stream) so a
|
||||
* misbehaving adapter cannot grow memory or corrupt a completed block.
|
||||
*/
|
||||
declare class BlockAssembler {
|
||||
/**
|
||||
* Feed one chunk into the assembly state.
|
||||
* @param chunk - the next raw chunk, in stream order.
|
||||
*/
|
||||
push(chunk: StreamChunk): void;
|
||||
/**
|
||||
* Assemble all blocks seen so far, in stream order.
|
||||
* @returns one block per seen index; an open block assembles from its
|
||||
* accumulated deltas (an unknown block type never closed by `block-end` throws).
|
||||
*/
|
||||
blocks(): ContentBlock[];
|
||||
/** Usage from the `usage` chunk; undefined until one arrives. */
|
||||
get usage(): TokenUsage | undefined;
|
||||
/** Finish reason from the `finish` chunk; `{kind: 'stop'}` when the stream ended without one. */
|
||||
get finish(): FinishReason;
|
||||
/** Adapter-private replay state from the terminal finish chunk, if any. */
|
||||
get replayState(): unknown;
|
||||
/**
|
||||
* The assembled assistant message.
|
||||
* @returns an assistant-role message over `blocks()` (same open-block assembly rules).
|
||||
*/
|
||||
message(): Message;
|
||||
}
|
||||
```
|
||||
|
||||
## seam
|
||||
|
||||
`LlmAdapter` 是提供方 seam:继承它、实现 `stream()`、通过 `ctx.llm.registerAdapter(models, adapter)` 注册。`block-start`/`block-end` 的 `index` 关联加上 assembler 意味着适配器只需发出格式正确的分片,块重组不是各适配器需要操心的事。消费方接口(`ctx.llm.stream()`)与 `llm/stream` waterfall(瀑布式事件)在 [architecture.md § Content blocks and streaming](../architecture.md#content-blocks-and-streaming-dsh-llm) 中描述。
|
||||
`LlmAdapter` 是提供方 seam:创建子类、实现 `stream()`,再用 `ctx.llm.registerAdapter(providers, adapter)` 注册一个适配器实例。`GenerateOptions.provider` 选择已注册适配器;`GenerateOptions.model` 会传给该适配器,无需在生命周期启动时注册。重复提供方路由会原子失败。可选的 `providerInfo()` 与异步 `listModels()` 方法为 `LlmService.listProviders()` / `listModels()` 提供分离的 selector 元数据。该目录仅供参考,不是请求白名单:适配器仍是权威,并可接受未列出的模型 id。单独的 `resolveModelContext()` 查询会暴露确切路由上对正确性敏感的容量信息,但不会让目录成员关系具有权威性;缺失表示元数据未知,而不是路由无效。适配器查找发生在 `llm/stream` waterfall 的终端 continuation,因此 listener 可以在查找前短路调用,或路由一个可变的一次性请求。`block-start` / `block-end` 的 `index` 关联与 assembler 共同意味着适配器只需 emit 格式正确的 chunk——block 重组不是每个适配器各自的问题。消费方 surface(`ctx.llm.stream()`)与 `llm/stream` waterfall 见 [architecture.md § 内容 block 与流式传输](../architecture.md#content-blocks-and-streaming-dsh-llm)。
|
||||
|
||||
```ts public-api
|
||||
/**
|
||||
* Provider-wire adapter for the harness message and stream vocabulary. Register implementations
|
||||
* with `ctx.llm.registerAdapter(providers, adapter)`. Every provider HTTP request must include
|
||||
* `attributionHeaders()`; prove that at the wire or library header-hook boundary. The hand-rolled
|
||||
* DeepSeek and pi-ai adapters intentionally exercise this contract through different internals.
|
||||
*/
|
||||
declare abstract class LlmAdapter {
|
||||
/**
|
||||
* Describe one provider route owned by this adapter.
|
||||
* @param provider - a route passed to `registerAdapter()` for this instance.
|
||||
* @returns detached display metadata whose id must equal `provider`.
|
||||
*/
|
||||
providerInfo(provider: string): LlmProviderInfo;
|
||||
/**
|
||||
* List models this adapter can currently advertise for one owned provider.
|
||||
* The result is advisory: an adapter may accept unlisted model ids, and
|
||||
* consumers must not turn absence into request rejection.
|
||||
* @param _provider - one provider route owned by this adapter.
|
||||
* @returns discoverable models in adapter-preferred order.
|
||||
*/
|
||||
listModels(_provider: string): Promise<readonly LlmModelInfo[]>;
|
||||
/**
|
||||
* Resolve context capacity for one model accepted by this adapter. Absence
|
||||
* means the adapter does not know the capacity, not that routing is invalid.
|
||||
* @param _provider - one provider route owned by this adapter.
|
||||
* @param _model - exact model id passed to {@link GenerateOptions.model}.
|
||||
* @returns provider-owned context metadata, or `undefined` when unavailable.
|
||||
*/
|
||||
resolveModelContext(
|
||||
_provider: string,
|
||||
_model: string,
|
||||
): Promise<LlmModelContext | undefined>;
|
||||
/**
|
||||
* Stream one model call as raw chunks. The only required method.
|
||||
* @param options - the fully-assembled request; implementations must honor `options.signal`.
|
||||
* @returns the chunk stream, obeying the adapter contract documented on `StreamChunk`.
|
||||
*/
|
||||
abstract stream(options: GenerateOptions): AsyncIterable<StreamChunk>;
|
||||
}
|
||||
```
|
||||
|
||||
`ContentBlockType`(`index` 关联块所携带的键集合)派生自 `ContentBlockMap`:
|
||||
|
||||
```ts type-equiv
|
||||
/**
|
||||
* Merge-extensible content blocks keyed by `type`. New core blocks must land
|
||||
* with adapter, UI, and compaction support.
|
||||
*/
|
||||
interface ContentBlockMap {
|
||||
'text': TextBlock
|
||||
'reasoning': ReasoningBlock
|
||||
|
||||
Reference in New Issue
Block a user