Turtle
aeb48cc5f3
A TUI run against an unreachable endpoint failed with only 'fetch failed':
undici wraps transport failures in a bare TypeError whose diagnosis lives
on .cause, and every diagnostic seam rendered only error.message. The
readline front door additionally rendered failed turns as pure silence.
- dsh-llm: new errorChain(value) renders the full cause chain and
AggregateError members with circular/hostile-coercion containment.
- llm-deepseek: pre-response transport failures throw LlmError('NETWORK')
naming the endpoint and chaining the fetch TypeError; aborts keep their
DOMException so the loop still classifies them as cancellation.
- agent-loop: durable turn/end error messages and logger warnings render
through errorChain; local renderThrown copies removed.
- ui-stdio: failure turn/end reasons now render ([turn failed <code>],
[turn aborted], [turn rejected], output-token-limit); startup-failure
logs use errorChain.
- ui-tui: agent/error notices and the startup-failure line use errorChain.
86 lines
6.9 KiB
Markdown
86 lines
6.9 KiB
Markdown
# @deepseek-ai/dsh-llm-deepseek
|
|
|
|
DeepSeek chat-completions adapter for the harness LLM seam: hand-rolled `fetch` + SSE translation from the official wire format (source of truth: the API docs — guides/thinking_mode, guides/tool_calls, api/create-chat-completion) into the `StreamChunk` protocol.
|
|
|
|
A second, library-backed implementation of the same seam exists in `@deepseek-ai/dsh-llm-pi-ai`. This package always owns the `deepseek` provider route; mounting a pi-ai profile with `provider: deepseek` in the same context throws `LlmError('DUPLICATE_ADAPTER')` by design.
|
|
|
|
The package root exposes the Cordis plugin contract and `DeepSeekAdapter`; wire serialization, SSE parsing, and chunk translation helpers are not part of that root contract.
|
|
|
|
## Config
|
|
|
|
```yaml
|
|
- id: llm-deepseek
|
|
name: '@deepseek-ai/dsh-llm-deepseek'
|
|
config:
|
|
apiKey: !!js process.env.DEEPSEEK_API_KEY # or rely on the env fallback
|
|
baseURL: !!js process.env.DEEPSEEK_BASE_URL # default: https://api.deepseek.com
|
|
thinking: enabled # optional; provider default is enabled
|
|
reasoningEffort: high # optional; high | max — omitted ⇒ not sent
|
|
models: # optional; defaults to V4 Flash and V4 Pro
|
|
- id: deepseek-v4-flash
|
|
name: DeepSeek V4 Flash
|
|
- id: private-reasoner
|
|
description: Company-hosted reasoning model
|
|
```
|
|
|
|
The plugin registers the single provider route `deepseek`. A request selects it with `provider: deepseek`; its `model` is passed through as the wire `model` string, so changing DeepSeek models does not require lifecycle-time registration. Omitting `models` advertises `deepseek-v4-flash` and `deepseek-v4-pro`; an explicit list replaces those defaults, while `models: []` advertises none. Catalog entries are exposed through `ctx.llm.listModels('deepseek')` for clients such as ACP editors, but remain advisory: unlisted model ids still pass through unchanged. An omitted entry name defaults to its id. Registering another adapter for `deepseek` throws `LlmError('DUPLICATE_ADAPTER')`.
|
|
|
|
`reasoningEffort` is **omitted by default** — when unset, the `reasoning_effort` wire field is not sent and the server applies its own default for the model. The only accepted values are `high` and `max` (DeepSeek's official effort levels). It is meaningful only with thinking enabled (the provider default).
|
|
|
|
`thinking`/`reasoningEffort` are adapter-level request defaults serialized as the official top-level `thinking: {type}` / `reasoning_effort` wire fields. They live in adapter config (not `GenerateOptions`) to keep the core vocabulary provider-neutral.
|
|
|
|
## App attribution
|
|
|
|
Every request carries the shared attribution header from dsh-llm's `attributionHeaders()` - the mandatory `User-Agent` baseline identifying the harness (see [dsh-llm § App attribution](../llm/README.md#app-attribution-attributionts)). Direct DeepSeek requests and OpenAI-compatible gateway requests get no provider-specific app-attribution headers under this adapter contract; OpenRouter app attribution is deferred to a future explicit OpenRouter adapter or mode.
|
|
|
|
## Wire-format notes (verified live + against the official docs)
|
|
|
|
- Streaming only (`stream_options.include_usage` always on). `usage` may arrive attached to the finish chunk or as a trailing usage-only chunk — the translator defers both to `[DONE]`, so `usage` always precedes `finish` and nothing follows `finish`.
|
|
- The first thinking-mode chunk carries `reasoning_content: ""` — handled (no spurious reasoning block).
|
|
- **Reasoning passback rule**: on assistant turns that carried tool calls, `reasoning_content` is serialized back in history (required by the API in thinking mode); on tool-call-free turns it is dropped (ignored anyway — saves tokens).
|
|
- Cache accounting: `cacheReadTokens` ← `prompt_cache_hit_tokens` / `prompt_tokens_details.cached_tokens`; DeepSeek reports no cache-write metric.
|
|
|
|
## Errors
|
|
|
|
Non-2xx responses throw `LlmError` with stable codes: `AUTH` (401/403), `RATE_LIMIT` (429), `CONTEXT_WINDOW_EXCEEDED` (a 400 whose provider code, type, or message identifies context overflow), `INVALID_REQUEST` (other 400s), `SERVER` (5xx), `HTTP_<status>` otherwise. A transport failure before any response (DNS, refused connection, TLS, proxy) throws `NETWORK` naming the configured endpoint and chaining fetch's `TypeError: fetch failed` as `cause`, so `errorChain` renders the underlying diagnosis; an abort keeps its `DOMException` so the loop classifies it as cancellation. Protocol violations throw `STREAM_CLOSED` (no `[DONE]`) or `MALFORMED_RESPONSE` (bad JSON payload). Unknown wire `finish_reason`s (e.g. `content_filter`, `insufficient_system_resource`) become `finish {kind: 'error', code: <REASON>}` chunks.
|
|
|
|
## Testing
|
|
|
|
Unit suites run against a local `node:http` mock SSE server (no network). Real-API coverage lives in `tests/adapter.e2e.ts` (`pnpm run test:e2e`, key-gated): V4 Flash + V4 Pro across thinking enabled/disabled and both official effort levels, including the thinking+tools round trip with reasoning passback.
|
|
|
|
## Model Experience
|
|
|
|
### DeepSeek request
|
|
|
|
#### What the model sees
|
|
|
|
The selected DeepSeek model receives the harness system prompt, message history, tool schemas, stop sequences, and call config without adapter-authored prompt prose. On a prior assistant turn with tool calls, its reasoning content is passed back as required; reasoning from tool-call-free turns is omitted.
|
|
|
|
#### Token effect
|
|
|
|
Provider tokenization governs exact input. Conditional reasoning passback increases tool-round-trip context, while dropping other reasoning avoids paying those tokens again; cache-read usage is reported when available.
|
|
|
|
#### KV Cache effect
|
|
|
|
An unchanged assembled prefix is eligible for DeepSeek cache reuse, which this adapter reports in usage. A model-route change or any upstream prompt, schema, prefix, or history change may prevent reuse from the first changed token; reasoning passback appends during tool round trips.
|
|
|
|
### DeepSeek response
|
|
|
|
#### What the model sees
|
|
|
|
Reasoning, text, and raw-string tool arguments are translated into harness chunks for the loop to log and assemble.
|
|
|
|
#### Token effect
|
|
|
|
Generated tokens follow provider thinking and effort settings plus the request's `maxTokens`; only loop-retained blocks affect later input.
|
|
|
|
#### KV Cache effect
|
|
|
|
Loop-retained response blocks append to the next request and preserve its earlier reusable prefix; dropped blocks have no later cache effect. Changing the provider or model selects a different cache domain.
|
|
|
|
## Known Limitations and Deferred Work
|
|
|
|
- **`tool_choice` is not mapped** — not part of the core vocabulary (MVP cut, shared with the pi-ai twin).
|
|
- **Requests use raw `fetch`, not `@cordisjs/plugin-http`** — no shared proxy/interception configuration; adoption is deferred until a second adapter wants it (`TODO(http)`).
|
|
- **Serialization flattens user and tool-result content to text blocks** — plugin-added block types are skipped, and empty tool output crosses the wire as the literal `(no output)`.
|