Next refactors

lib/core/constants/ai_constants.dart

@@ -0,0 +1,49 @@
+import 'dart:io';
+
+class AiConstants {
+  AiConstants._();
+
+  // Server ports
+  static const int chatServerPort = 8080;
+  static const int embeddingServerPort = 8081;
+
+  // API endpoints
+  static String get chatApiUrl =>
+      'http://localhost:$chatServerPort/v1/chat/completions';
+  static String get embeddingApiUrl =>
+      'http://localhost:$embeddingServerPort/v1/embeddings';
+
+  // Model files
+  static const String qwenModelFile = 'qwen2.5-7b-instruct-q4_k_m.gguf';
+  static const String nomicModelFile = 'nomic-embed-text-v1.5.Q4_K_M.gguf';
+  static const String nomicModelName = 'nomic-embed-text-v1.5.Q4_K_M';
+
+  // llama.cpp binary
+  static const String llamaBuild = 'b8130';
+  static String get serverBinaryName =>
+      Platform.isWindows ? 'llama-server.exe' : 'llama-server';
+
+  // Model download URLs
+  static const String nomicModelUrl =
+      'https://huggingface.co/nomic-ai/nomic-embed-text-v1.5-GGUF/resolve/main/nomic-embed-text-v1.5.Q4_K_M.gguf';
+  static const String qwenModelUrl =
+      'https://huggingface.co/bartowski/Qwen2.5-7B-Instruct-GGUF/resolve/main/Qwen2.5-7B-Instruct-Q4_K_M.gguf';
+
+  // Server configuration
+  static const int qwenContextSize = 4096;
+  static const int nomicContextSize = 8192;
+  static const int gpuLayerOffload = 99;
+  static const double chatTemperature = 0.7;
+
+  // Timeouts
+  static const Duration serverConnectTimeout = Duration(seconds: 30);
+  static const Duration serverReceiveTimeout = Duration(minutes: 5);
+  static const Duration embeddingConnectTimeout = Duration(seconds: 10);
+  static const Duration embeddingReceiveTimeout = Duration(seconds: 60);
+
+  // System prompt
+  static const String baseSystemPrompt =
+      'You are a helpful AI fitness assistant for personal trainers. '
+      'Help users design training plans, analyse exercise technique, '
+      'and answer questions about sports science and nutrition.';
+}