Coder/deepseek-harness
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

refactor(llm): simplify live request telemetry (round 2)

Browse Source
This commit is contained in:
Hypatia May
2026-07-28 19:03:45 +08:00
parent 35b9c454e5
commit e6ce6abd7d
31 changed files with 159 additions and 187 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
docs/core-data-structures/llm-streaming.md
View File

@@ -161,7 +161,7 @@ declare class BlockAssembler {
## The seam
`LlmAdapter` is the provider seam: subclass, implement `stream()`, and register one adapter instance with `ctx.llm.registerAdapter(providers, adapter)`. `GenerateOptions.provider` selects the registered adapter; `GenerateOptions.model` is passed to that adapter and need not be registered at lifecycle start. Duplicate provider routes fail atomically. Optional `providerRetryPolicy()` is captured per route with normal defaults, while `providerInfo()` and asynchronous `listModels()` feed `LlmService.listProviders()` / `listModels()` with detached selector metadata. That catalog is advisory rather than a request whitelist: the adapter remains authoritative and may accept unlisted model ids. One asynchronous `resolveModel()` query returns exact model identity plus optional correctness-sensitive context capacity and ordered model-owned reasoning ids with an optional deployment default; absent fields mean unavailable metadata or capability, not invalid catalog membership. The resolver receives optional cancellation and must settle promptly after abort. `LlmService.resolveModelInfo()` validates and detaches the aggregate. The service validates and materializes reasoning through `resolveCallConfig()` at the final adapter boundary, so direct calls cannot bypass unsupported-effort rejection; direct dispatch captures one registration before awaiting that resolution. The agent loop instead uses `prepareCall()` to keep the same registration across model resolution, durable header logging, and dispatch, and to retain detached context metadata from that exact lookup. Its optional observer runs after a final stream handle is constructed and before adapter iteration. Adapter lookup happens at the terminal continuation of the `llm/stream` waterfall, so a listener may short-circuit the call or route a mutable one-shot request before lookup. The `block-start` / `block-end` `index` correlation and the assembler together mean an adapter only has to emit well-formed chunks — block reassembly is not each adapter's problem. The consumer surface (`ctx.llm.stream()`) and the `llm/stream` waterfall are described in [architecture.md § Content blocks and streaming](../architecture.md#content-blocks-and-streaming-dsh-llm).
`LlmAdapter` is the provider seam: subclass, implement `stream()`, and register one adapter instance with `ctx.llm.registerAdapter(providers, adapter)`. `GenerateOptions.provider` selects the registered adapter; `GenerateOptions.model` is passed to that adapter and need not be registered at lifecycle start. Duplicate provider routes fail atomically. Optional `providerRetryPolicy()` is captured per route with normal defaults, while `providerInfo()` and asynchronous `listModels()` feed `LlmService.listProviders()` / `listModels()` with detached selector metadata. That catalog is advisory rather than a request whitelist: the adapter remains authoritative and may accept unlisted model ids. One asynchronous `resolveModel()` query returns exact model identity plus optional correctness-sensitive context capacity and ordered model-owned reasoning ids with an optional deployment default; absent fields mean unavailable metadata or capability, not invalid catalog membership. The resolver receives optional cancellation and must settle promptly after abort. `LlmService.resolveModelInfo()` validates and detaches the aggregate. The service validates and materializes reasoning through `resolveCallConfig()` at the final adapter boundary, so direct calls cannot bypass unsupported-effort rejection; direct dispatch captures one registration before awaiting that resolution. The agent loop instead uses `prepareCall()` to keep the same registration across model resolution, durable header logging, and dispatch, and to retain detached context metadata from that exact lookup. Adapter lookup happens at the terminal continuation of the `llm/stream` waterfall, so a listener may short-circuit the call or route a mutable one-shot request before lookup. AgentLoop observes a request attempt once the outer waterfall returns a stream handle; that limited boundary does not prove a lazy terminal adapter was constructed or began provider I/O. The `block-start` / `block-end` `index` correlation and the assembler together mean an adapter only has to emit well-formed chunks — block reassembly is not each adapter's problem. The consumer surface (`ctx.llm.stream()`) and the `llm/stream` waterfall are described in [architecture.md § Content blocks and streaming](../architecture.md#content-blocks-and-streaming-dsh-llm).
```ts type-equiv
/** One model call whose config and adapter registration were resolved together. */
@@ -175,11 +175,9 @@ interface PreparedLlmCall {
* preparation. The request's call-config fields must match {@link config};
* reuse or mismatch fails with `INVALID_PREPARED_CALL`.
* @param options - fully assembled request carrying the prepared config.
* @param onDispatched - contained Agent-loop notification hook invoked after
* a stream handle is constructed and before its adapter is iterated.
* @returns the chunk stream, including the `llm/stream` waterfall.
*/
stream(options: GenerateOptions, onDispatched?: () => void): AsyncIterable<StreamChunk>
stream(options: GenerateOptions): AsyncIterable<StreamChunk>
}
```