feat: externalize AI configuration, implement resilience policies, and update extraction prompt formatting

src/NexusReader.Infrastructure/Services/KnowledgeService.cs

@@ -8,7 +8,9 @@ using NexusReader.Domain.Entities;
 using NexusReader.Infrastructure.Helpers;
 using NexusReader.Infrastructure.Persistence;
 using Polly;
-using Polly.Retry;
+using Polly.Registry;
+using Microsoft.Extensions.Options;
+using NexusReader.Infrastructure.Configuration;
 
 namespace NexusReader.Infrastructure.Services;
 
@@ -16,25 +18,20 @@ public class KnowledgeService : IKnowledgeService
 {
     private readonly IChatClient _chatClient;
     private readonly AppDbContext _dbContext;
+    private readonly ResiliencePipeline _retryPipeline;
+    private readonly AiSettings _settings;
     private const string PromptVersion = "1.0";
-    private const string ModelId = "gemini-1.5-flash";
 
-    private static readonly ResiliencePipeline _retryPipeline = new ResiliencePipelineBuilder()
-        .AddRetry(new RetryStrategyOptions
-        {
-            ShouldHandle = new PredicateBuilder().Handle<Exception>(ex => 
-                ex.Message.Contains("429") || ex.Message.Contains("Too Many Requests") || ex.Message.Contains("quota")),
-            BackoffType = DelayBackoffType.Exponential,
-            UseJitter = true,
-            MaxRetryAttempts = 3,
-            Delay = TimeSpan.FromSeconds(2)
-        })
-        .Build();
-
-    public KnowledgeService(IChatClient chatClient, AppDbContext dbContext)
+    public KnowledgeService(
+        IChatClient chatClient, 
+        AppDbContext dbContext,
+        ResiliencePipelineProvider<string> pipelineProvider,
+        IOptions<AiSettings> settings)
     {
         _chatClient = chatClient;
         _dbContext = dbContext;
+        _retryPipeline = pipelineProvider.GetPipeline("ai-retry");
+        _settings = settings.Value;
     }
 
     public async Task<Result<KnowledgePacket>> GetKnowledgeAsync(string text, CancellationToken cancellationToken = default)
@@ -48,10 +45,9 @@ public class KnowledgeService : IKnowledgeService
         var normalizedText = ContentHasher.Normalize(text);
 
         // Phase 4: Request Pre-processing (Token Saving)
-        const int MaxInputLength = 15000; // Roughly 3k-4k tokens
-        if (normalizedText.Length > MaxInputLength)
+        if (normalizedText.Length > _settings.MaxInputLength)
         {
-            return Result.Fail($"Input text is too long ({normalizedText.Length} characters after normalization). Max allowed is {MaxInputLength}.");
+            return Result.Fail($"Input text is too long ({normalizedText.Length} characters after normalization). Max allowed is {_settings.MaxInputLength}.");
         }
 
         // Simple token estimation (4 chars per token)
@@ -86,8 +82,8 @@ public class KnowledgeService : IKnowledgeService
             var options = new ChatOptions
             {
                 ResponseFormat = ChatResponseFormat.Json,
-                Temperature = 0.1f,
-                MaxOutputTokens = 1000
+                Temperature = (float)_settings.Temperature,
+                MaxOutputTokens = _settings.MaxOutputTokens
             };
 
             var response = await _retryPipeline.ExecuteAsync(async ct => 
@@ -117,7 +113,7 @@ public class KnowledgeService : IKnowledgeService
             {
                 ContentHash = hash,
                 JsonData = jsonResponse,
-                ModelId = ModelId,
+                ModelId = _settings.Model,
                 PromptVersion = PromptVersion,
                 CreatedAt = DateTime.UtcNow
             };