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

internal/gateway/chat.go

@@ -651,6 +651,15 @@ func mergeUsage(prev, cur *types.TokenUsage) *types.TokenUsage {
 		out.Completion = cur.Completion
 	}
 	out.Total = out.Prompt + out.Completion
+	if cur.PromptTokensDetails != nil && cur.PromptTokensDetails.CachedTokens > 0 {
+		out.PromptTokensDetails = cur.PromptTokensDetails
+	}
+	if cur.PromptCacheHit > 0 {
+		out.PromptCacheHit = cur.PromptCacheHit
+	}
+	if cur.PromptCacheMiss > 0 {
+		out.PromptCacheMiss = cur.PromptCacheMiss
+	}
 	return &out
 }