fix: 将诊断日志级别提升为 WARN 以确保输出

将所有诊断日志从 DEBUG 改为 WARN 级别，
确保在测试运行时能够捕获到这些关键信息。

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Sonnet 4.5 <[email protected]>

Author: CodeCasterX

framework/fel/java/fel-flow/src/main/java/modelengine/fel/engine/activities/AiStart.java

@@ -596,18 +596,18 @@ public final AiState<Tip, D, Tip, RF, F> runnableParallel(Pattern<O, Tip>... pat
 
         AiState<Tip, D, Tip, RF, F> state = aiFork.join(Tip::new, (acc, data) -> {
             // 诊断日志：记录每次reducer调用的详细信息
-            log.debug("[Fork.join reducer] Thread={}, acc={}, data={}, data_is_null={}",
+            log.warn("[DIAGNOSTIC Fork.join reducer] Thread={}, acc={}, data={}, data_is_null={}",
                     Thread.currentThread().getName(),
                     acc,
                     data,
                     data == null);
 
             // Tip.merge() 内部会处理 data 为 null 的情况
             if (data == null) {
-                log.warn("[Fork.join reducer] DIAGNOSTIC: Received null data in reducer! acc={}, thread={}",
+                log.warn("[DIAGNOSTIC Fork.join reducer] Received null data in reducer! acc={}, thread={}",
                         acc, Thread.currentThread().getName());
                 // 打印堆栈跟踪以了解调用路径
-                log.warn("[Fork.join reducer] Stack trace:", new RuntimeException("null data diagnostic"));
+                log.warn("[DIAGNOSTIC Fork.join reducer] Stack trace:", new RuntimeException("null data diagnostic"));
             }
             acc.merge(data);
             return acc;
@@ -620,20 +620,20 @@ private Processor<O, Tip> getPatternProcessor(Pattern<O, Tip> pattern, AiState<O
         return node.publisher()
                 .map(input -> {
                     O inputData = input.getData();
-                    log.debug("[getPatternProcessor] Executing pattern={}, inputData={}, thread={}",
+                    log.warn("[DIAGNOSTIC getPatternProcessor] Executing pattern={}, inputData={}, thread={}",
                             pattern.getClass().getSimpleName(),
                             inputData,
                             Thread.currentThread().getName());
 
                     Tip result = AiFlowSession.applyPattern(pattern, inputData, input.getSession());
 
-                    log.debug("[getPatternProcessor] Pattern result={}, result_is_null={}, thread={}",
+                    log.warn("[DIAGNOSTIC getPatternProcessor] Pattern result={}, result_is_null={}, thread={}",
                             result,
                             result == null,
                             Thread.currentThread().getName());
 
                     if (result == null) {
-                        log.error("[getPatternProcessor] CRITICAL: Pattern returned null! pattern={}, inputData={}",
+                        log.error("[DIAGNOSTIC getPatternProcessor] CRITICAL: Pattern returned null! pattern={}, inputData={}",
                                 pattern.getClass().getSimpleName(),
                                 inputData);
                     }