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

Merge remote-tracking branch 'origin/codex/package-readme-limitations-audit-20260712' into codex/model-experience-readmes-20260712

Browse Source 
# Conflicts:
#	docs/cookbook/adding-a-package.md
#	docs/rfc/INDEX.md
#	package.json
#	packages/AGENTS.md
#	packages/bash/bash-sandbox/README.md
#	packages/sandbox/sandbox-local/README.md
#	packages/sandbox/sandbox/README.md
#	packages/session-persistence/session-persistence-sqlite/README.md
#	packages/support/acp-snapshot/README.md
#	packages/ui/app-boot/README.md
#	packages/ui/user-approval/README.md
#	packages/workflow/tool-workflow/README.md
This commit is contained in:
Tianyi Cui
2026-07-12 02:48:49 +08:00
parent 027970043c 41a7c70cb1
commit a08485fc80
73 changed files with 623 additions and 58 deletions
Download Patch File Download Diff File Expand all files Collapse all files
packages/llm/llm-deepseek/README.md
+6 -4
View File
@@ -41,10 +41,6 @@ Every request carries the shared attribution header from dsh-llm's `attributionH
- **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.
## Limitations (MVP, documented deliberately)
- `tool_choice` is not mapped (not part of the core vocabulary).
## Errors
Non-2xx responses throw `LlmError` with stable codes: `AUTH` (401/403), `RATE_LIMIT` (429), `INVALID_REQUEST` (400), `SERVER` (5xx), `HTTP_<status>` otherwise. 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.
@@ -52,3 +48,9 @@ Non-2xx responses throw `LlmError` with stable codes: `AUTH` (401/403), `RATE_LI
## 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.
## 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)`.