dev 045ecf47bc feat(types): pass through upstream cache tokens in TokenUsage ...

dsh displays cache-hit %, but llmsproxy dropped every upstream's cache
fields — deepseek prompt_cache_hit_tokens, OpenAI prompt_tokens_details.
cached_tokens, anthropic cache_read_input_tokens, gemini cachedContentTokenCount.

Changes:
- TokenUsage: add PromptTokensDetails (with CachedTokens) + PromptCacheHit/Miss
- MarshalJSON: emit prompt_tokens_details.cached_tokens (OpenAI v2 standard)
  and prompt_cache_hit/miss_tokens (DeepSeek legacy) — dsh reads the former
  first, falls back to the latter
- mergeUsage: preserve cache fields across stream chunks
- standardSSEChunk: parse the upstream raw prompt_tokens_details too
- deepseek.lua: forward prompt_cache_hit/miss_tokens + create
  prompt_tokens_details from them
- openai.lua: forward prompt_tokens_details.cached_tokens and legacy
  prompt_cache_hit/miss_tokens; normalize legacy hits into the standard
  object so dsh sees them regardless of upstream format
- anthropic.lua: map cache_read_input_tokens → prompt_tokens_details
- gemini.lua: map cachedContentTokenCount → prompt_tokens_details

2026-08-25 07:57:32 +08:00

..

adapters

feat(types): pass through upstream cache tokens in TokenUsage

2026-08-25 07:57:32 +08:00

vm_test.go

feat(gateway): pass through upstream finish_reason end-to-end

2026-08-24 15:05:50 +08:00

vm.go

feat(adapter): move per-source error condensing into transform_error hooks

2026-08-24 19:17:36 +08:00