Update llm prompt sys

src/Managing.Api/Controllers/LlmController.cs

@@ -1,3 +1,4 @@
+using System.Text.Json;
 using Managing.Application.Abstractions.Services;
 using Microsoft.AspNetCore.Authorization;
 using Microsoft.AspNetCore.Mvc;
@@ -125,7 +126,7 @@ public class LlmController : BaseController
                         return new LlmMessage
                         {
                             Role = "tool",
-                            Content = System.Text.Json.JsonSerializer.Serialize(toolResult),
+                            Content = JsonSerializer.Serialize(toolResult),
                             ToolCallId = toolCall.Id
                         };
                     }
@@ -223,10 +224,21 @@ public class LlmController : BaseController
     {
         var lastMessage = request.Messages.LastOrDefault(m => m.Role == "user")?.Content?.ToLowerInvariant() ?? "";
 
-        // Complex operations need more iterations
+        // Complex operations need more iterations (bundle analysis, multi-step workflows)
-        if (lastMessage.Contains("bundle") || lastMessage.Contains("analyze") || lastMessage.Contains("compare"))
+        if (lastMessage.Contains("bundle") || lastMessage.Contains("compare") || lastMessage.Contains("all backtests"))
             return 5;
 
+        // Backtest detail requests need 4 iterations (list → get_by_id → analyze → format)
+        if (lastMessage.Contains("backtest") &&
+            (lastMessage.Contains("detail") || lastMessage.Contains("analyze") || lastMessage.Contains("show") ||
+             lastMessage.Contains("this") || lastMessage.Contains("that") || lastMessage.Contains("best") ||
+             lastMessage.Contains("top") || lastMessage.Contains("recent")))
+            return 4;
+
+        // General analysis queries
+        if (lastMessage.Contains("analyze"))
+            return 4;
+
         // Simple queries need fewer iterations
         if (lastMessage.Contains("explain") || lastMessage.Contains("what is") || lastMessage.Contains("how does"))
             return 2;
@@ -264,20 +276,34 @@ public class LlmController : BaseController
                - Indicators: get_indicator_info() for detailed specs
                - Use conversation context: "that X" or "this Y" → extract ID from previous messages
 
-            2. ANALYZE WITH EXPERTISE:
+            2. CONTEXT EXTRACTION:
+               - Pay attention to backtest IDs mentioned in conversation history
+               - When user says "analyze this backtest" or "show me details", extract the backtest ID from previous messages
+               - If multiple backtests were listed, use the most recently mentioned one or the top-ranked one
+               - NEVER ask user for IDs that were already provided in conversation
+
+            3. BACKTEST DETAIL WORKFLOW:
+               When user requests backtest details/analysis:
+               a) If backtest ID is in conversation → IMMEDIATELY call get_backtest_by_id(id)
+               b) If no ID but refers to "best/top" → call get_backtests_paginated(sortBy='Score', sortOrder='desc', pageSize=1) THEN get_backtest_by_id()
+               c) If no ID but refers to "recent/latest" → call get_backtests_paginated(sortOrder='desc', pageSize=1) THEN get_backtest_by_id()
+               d) If completely ambiguous → ask ONCE for clarification, then proceed
+
+            4. ANALYZE WITH EXPERTISE:
                After retrieving data, provide comprehensive analysis:
 
-               Backtests: Performance (PnL, growth, ROI), Risk (Sharpe, drawdown), Win rate, Position patterns, Strengths/weaknesses, Recommendations
+               Backtests: Performance (PnL, growth, ROI), Risk (Sharpe, drawdown), Win rate, Position patterns, Trade duration, Strengths/weaknesses, Recommendations
                Bundles: Aggregate performance, Best/worst combinations, Optimal parameters, Robustness
                Indicators: Use cases, Parameter sensitivity, Combination suggestions, Pitfalls
                General: Compare to benchmarks, Statistical significance, Actionable insights
 
-            3. BE PROACTIVE:
+            5. BE PROACTIVE:
                - Execute multiple tool iterations for complete data
                - Interpret data, don't just return it
+               - Chain tool calls automatically (list → get_by_id → analyze)
                - Only ask for clarification when truly ambiguous
 
-            Be concise, accurate, and proactive.
+            Be concise, accurate, and proactive. Always prioritize retrieving complete data over asking questions.
             """;
     }