refactor: remove debug prints and add logging for todo updates

src/main/java/io/mybatis/learn/s03/S03TodoWrite.java

@@ -85,11 +85,6 @@ public class S03TodoWrite implements CommandLineRunner {
                     .call()
                     .content();
 
-            // 每次回复后打印当前 todo 状态
-            System.out.println("\n--- Todo Status ---");
-            System.out.println(todoManager.render());
-            System.out.println("-------------------");
-
             return response;
         });
     }

src/main/java/io/mybatis/learn/s03/TodoManager.java

@@ -1,5 +1,7 @@
 package io.mybatis.learn.s03;
 
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
 import org.springframework.ai.tool.annotation.Tool;
 import org.springframework.ai.tool.annotation.ToolParam;
 
@@ -23,6 +25,8 @@ import java.util.List;
  */
 public class TodoManager {
 
+    private static final Logger log = LoggerFactory.getLogger(TodoManager.class);
+
     /**
      * Todo 项数据结构。
      * TIP: 对应 Python 中 todo item 字典 {@code {"id": ..., "text": ..., "status": ...}}。
@@ -73,7 +77,11 @@ public class TodoManager {
         }
 
         this.items = validated;
-        return render();
+        String rendered = render();
+        if (log.isDebugEnabled()) {
+            System.out.printf("📋 调用工具 [Todo] 更新 %d 项任务%n%s%n", items.size(), rendered);
+        }
+        return rendered;
     }
 
     /**

src/main/java/io/mybatis/learn/s08/BackgroundManager.java

@@ -24,7 +24,13 @@ import java.util.stream.Collectors;
  * <p>
  * TIP: 对应 Python {@code agents/s08_background_tasks.py} 中的 {@code BackgroundManager} 类。
  * Python 使用 {@code threading.Thread(daemon=True)}，
- * Java 使用 {@link ExecutorService} + 虚拟线程（Java 21）。
+ * Java 同样使用 daemon 平台线程（而非虚拟线程）。
+ * <p>
+ * ⚠️ 为何不用虚拟线程：{@code execute()} 内通过 {@code BufferedReader.lines()} 读取进程输出，
+ * 底层调用 {@code FileInputStream.read0()}（native 方法），会将虚拟线程
+ * <b>钉住（pin）在载体线程（carrier thread）</b> 上。
+ * 载体线程池大小 = {@code availableProcessors()}，若池耗尽则后续任务只能串行等待。
+ * daemon 平台线程没有此约束，可真正并行执行多个阻塞 I/O 任务，符合 Python 版行为。
  * <pre>
  *   Main thread              Background thread
  *   +-----------------+      +-----------------+
@@ -42,7 +48,14 @@ public class BackgroundManager {
 
     private final Map<String, TaskInfo> tasks = new ConcurrentHashMap<>();
     private final List<Notification> notificationQueue = new CopyOnWriteArrayList<>();
-    private final ExecutorService executor = Executors.newVirtualThreadPerTaskExecutor();
+    // TIP: 使用 daemon 平台线程而非虚拟线程。
+    // 读取进程输出时底层会调用 native read0()，虚拟线程遇到 native 方法会被 pin 在载体线程上，
+    // 导致多个后台任务串行排队。平台线程无此限制，对应 Python 的 threading.Thread(daemon=True)。
+    private final ExecutorService executor = Executors.newCachedThreadPool(r -> {
+        Thread t = new Thread(r, "bg-worker");
+        t.setDaemon(true);
+        return t;
+    });
     private final String workDir;
 
     record TaskInfo(String status, String result, String command) {
@@ -56,7 +69,9 @@ public class BackgroundManager {
         log.info("BackgroundManager 初始化，workDir={}", workDir);
     }
 
-    @Tool(description = "Run a command in a background thread. Returns task_id immediately without waiting.")
+    @Tool(description = "Run a command in a background thread. Returns task_id immediately without waiting. "
+            + "When starting multiple independent background tasks, call this tool for ALL of them "
+            + "in a single response (parallel function calls) so they start at the same time.")
     public String backgroundRun(
             @ToolParam(description = "The shell command to run in background") String command) {
         String taskId = UUID.randomUUID().toString().substring(0, 8);

src/main/java/io/mybatis/learn/s08/S08BackgroundTasks.java

@@ -50,8 +50,15 @@ public class S08BackgroundTasks implements CommandLineRunner {
                 System.out.println("[Background tasks completed: " + notifs.size() + "]");
             }
 
+            // TIP: 明确要求 LLM 并行调用 backgroundRun，避免逐个调用导致任务串行提交。
+            // 同时注入当前 OS 信息，让 LLM 选择平台适配的命令（如 Windows 用 timeout /t N /nobreak 替代 sleep N）。
+            String os = System.getProperty("os.name");
             String system = "You are a coding agent at " + workDir
-                    + ". Use backgroundRun for long-running commands." + bgContext;
+                    + ". Current OS: " + os + "."
+                    + " Use backgroundRun for long-running commands."
+                    + " When starting multiple independent background tasks, issue ALL backgroundRun"
+                    + " calls in a single parallel batch so they begin simultaneously."
+                    + bgContext;
 
             ChatClient chatClient = ChatClient.builder(chatModel)
                     .defaultSystem(system)