fix: extract and log Claude output_config.effort in usage records

Claude's output_config.effort parameter (low/medium/high/max) was not
being extracted from requests or logged in the reasoning_effort column
of usage logs. Only the OpenAI path populated this field.

Changes:
- Extract output_config.effort in ParseGatewayRequest
- Add ReasoningEffort field to ForwardResult
- Populate reasoning_effort in both RecordUsage and RecordUsageWithLongContext
- Guard against overwriting service-set effort values in handler
- Update stale comments that described reasoning_effort as OpenAI-only
- Add unit tests for extraction, normalization, and persistence

backend/internal/service/usage_log.go

@@ -100,8 +100,9 @@ type UsageLog struct {
 	Model     string
 	// ServiceTier records the OpenAI service tier used for billing, e.g. "priority" / "flex".
 	ServiceTier *string
-	// ReasoningEffort is the request's reasoning effort level (OpenAI Responses API),
-	// e.g. "low" / "medium" / "high" / "xhigh". Nil means not provided / not applicable.
+	// ReasoningEffort is the request's reasoning effort level.
+	// OpenAI: "low" / "medium" / "high" / "xhigh"; Claude: "low" / "medium" / "high" / "max".
+	// Nil means not provided / not applicable.
 	ReasoningEffort *string
 
 	GroupID        *int64