fix(distill): increase HTTP client timeout 30s→120s for LLM calls

go/internal/distill/engine.go

@@ -120,7 +120,7 @@ func NewEngine(llmEndpoint, llmModel, apiKey string) *Engine {
 		batchSize:     10,
 		batchTimeout:  5 * time.Minute,
 		dailyLimit:    5000,
-		client:        &http.Client{Timeout: 30 * time.Second},
+		client:        &http.Client{Timeout: 120 * time.Second},
 	}
 }