diff --git a/domain/processors/builtin/vector_index.py b/domain/processors/builtin/vector_index.py
index edc9e0c..3a7fb26 100644
--- a/domain/processors/builtin/vector_index.py
+++ b/domain/processors/builtin/vector_index.py
@@ -114,8 +114,15 @@ class VectorIndexProcessor:
         }
     ]
     produces_file = False
+    requires_input_bytes = False
 
     async def process(self, input_bytes: bytes, path: str, config: Dict[str, Any]) -> Response:
+        async def ensure_input_bytes() -> bytes:
+            if input_bytes:
+                return input_bytes
+            from domain.virtual_fs import VirtualFSService
+            return await VirtualFSService.read_file(path)
+
         action = config.get("action", "create")
         index_type = config.get("index_type", "vector")
         vector_db = VectorDBService()
@@ -159,7 +166,8 @@ class VectorIndexProcessor:
         await vector_db.delete_vector(vector_collection, path)
 
         if file_ext in ["jpg", "jpeg", "png", "bmp"]:
-            processed_bytes, compression = _compress_image_for_embedding(input_bytes)
+            file_bytes = await ensure_input_bytes()
+            processed_bytes, compression = _compress_image_for_embedding(file_bytes)
             base64_image = base64.b64encode(processed_bytes).decode("utf-8")
             description = await describe_image_base64(base64_image)
             embedding = await get_text_embedding(description)
@@ -180,7 +188,8 @@ class VectorIndexProcessor:
 
         if file_ext in ["txt", "md"]:
             try:
-                text = input_bytes.decode("utf-8")
+                file_bytes = await ensure_input_bytes()
+                text = file_bytes.decode("utf-8")
             except UnicodeDecodeError:
                 return Response(content="文本文件解码失败", status_code=400)
 
diff --git a/domain/processors/service.py b/domain/processors/service.py
index 2305496..a3061b7 100644
--- a/domain/processors/service.py
+++ b/domain/processors/service.py
@@ -85,6 +85,44 @@ class ProcessorService:
         suffix = raw_suffix
         overwrite = req.overwrite
 
+        if produces_file:
+            if not overwrite and not suffix:
+                raise HTTPException(400, detail="Suffix is required when not overwriting files")
+        else:
+            overwrite = False
+            suffix = None
+        payload = {
+            "path": req.path,
+            "processor_type": req.processor_type,
+            "config": req.config,
+            "overwrite": overwrite,
+            "max_depth": req.max_depth,
+            "suffix": suffix,
+        }
+        task = await task_queue_service.add_task("process_directory_scan", payload)
+        return {"task_id": task.id}
+
+    @classmethod
+    async def scan_directory(cls, req: ProcessDirectoryRequest):
+        if req.max_depth is not None and req.max_depth < 0:
+            raise HTTPException(400, detail="max_depth must be >= 0")
+
+        is_dir = await VirtualFSService.path_is_directory(req.path)
+        if not is_dir:
+            raise HTTPException(400, detail="Path must be a directory")
+
+        schema = get_config_schema(req.processor_type)
+        _processor = get(req.processor_type)
+        if not schema or not _processor:
+            raise HTTPException(404, detail="Processor not found")
+
+        produces_file = bool(schema.get("produces_file"))
+        raw_suffix = req.suffix if req.suffix is not None else None
+        if raw_suffix is not None and raw_suffix.strip() == "":
+            raw_suffix = None
+        suffix = raw_suffix
+        overwrite = req.overwrite
+
         if produces_file:
             if not overwrite and not suffix:
                 raise HTTPException(400, detail="Suffix is required when not overwriting files")
@@ -133,7 +171,7 @@ class ProcessorService:
                 new_name = f"{name}{suffix_str}"
             return str(path_obj.with_name(new_name))
 
-        scheduled_tasks: List[str] = []
+        scheduled_count = 0
         stack: List[Tuple[str, int]] = [(rel, 0)]
         page_size = 200
 
@@ -161,7 +199,7 @@ class ProcessorService:
                     save_to = None
                     if produces_file and not overwrite and suffix:
                         save_to = apply_suffix(absolute_path, suffix)
-                    task = await task_queue_service.add_task(
+                    await task_queue_service.add_task(
                         "process_file",
                         {
                             "path": absolute_path,
@@ -171,16 +209,13 @@ class ProcessorService:
                             "overwrite": overwrite,
                         },
                     )
-                    scheduled_tasks.append(task.id)
+                    scheduled_count += 1
 
                 if total is None or page * page_size >= total:
                     break
                 page += 1
 
-        return {
-            "task_ids": scheduled_tasks,
-            "scheduled": len(scheduled_tasks),
-        }
+        return {"scheduled": scheduled_count}
 
     @classmethod
     async def get_source(cls, processor_type: str):
diff --git a/domain/tasks/task_queue.py b/domain/tasks/task_queue.py
index e9447f4..9d36886 100644
--- a/domain/tasks/task_queue.py
+++ b/domain/tasks/task_queue.py
@@ -86,6 +86,12 @@ class TaskQueueService:
                     overwrite=params.get("overwrite", False),
                 )
                 task.result = result
+            elif task.name == "process_directory_scan":
+                from domain.processors import ProcessDirectoryRequest, ProcessorService
+
+                params = task.task_info or {}
+                req = ProcessDirectoryRequest(**params)
+                task.result = await ProcessorService.scan_directory(req)
             elif task.name == "automation_task" or self._is_processor_task(task.name):
                 from models.database import AutomationTask
 
diff --git a/web/src/api/processors.ts b/web/src/api/processors.ts
index 578cf2d..32bddc6 100644
--- a/web/src/api/processors.ts
+++ b/web/src/api/processors.ts
@@ -43,7 +43,7 @@ export const processorsApi = {
     max_depth?: number | null;
     suffix?: string | null;
   }) =>
-    request<{ task_ids: string[]; scheduled: number }>('/processors/process-directory', {
+    request<{ task_id: string }>('/processors/process-directory', {
       method: 'POST',
       json: params,
     }),
diff --git a/web/src/i18n/locales/en.json b/web/src/i18n/locales/en.json
index fd2e3be..f409915 100644
--- a/web/src/i18n/locales/en.json
+++ b/web/src/i18n/locales/en.json
@@ -617,6 +617,7 @@
   "Source Editor": "Source Editor",
   "Module Path": "Module Path",
   "Directory processing always overwrites original files": "Directory processing always overwrites original files",
+  "Directory execution will enqueue one task per file": "Directory execution will enqueue a scan task, then one task per file",
   "No data": "No data",
   "Select File": "Select File",
   "Select Path": "Select Path",
diff --git a/web/src/i18n/locales/zh.json b/web/src/i18n/locales/zh.json
index c87acaf..357888f 100644
--- a/web/src/i18n/locales/zh.json
+++ b/web/src/i18n/locales/zh.json
@@ -608,7 +608,7 @@
   "Source Editor": "源码编辑",
   "Module Path": "模块路径",
   "Directory processing always overwrites original files": "选择目录时会强制覆盖原文件",
-  "Directory execution will enqueue one task per file": "目录模式会为每个文件单独创建任务",
+  "Directory execution will enqueue one task per file": "目录模式会先创建扫描任务，后台再为每个文件创建任务",
   "Directory scope": "目录范围",
   "Current level only": "仅当前层级",
   "Include subdirectories": "包含子目录",
diff --git a/web/src/pages/ProcessorsPage.tsx b/web/src/pages/ProcessorsPage.tsx
index 2c18532..a7c9091 100644
--- a/web/src/pages/ProcessorsPage.tsx
+++ b/web/src/pages/ProcessorsPage.tsx
@@ -276,7 +276,7 @@ const ProcessorsPage = memo(function ProcessorsPage() {
             max_depth: maxDepth,
             suffix: suffixValue,
           });
-          messageApi.success(`${t('Task submitted')}: ${resp.scheduled}`);
+          messageApi.success(`${t('Task submitted')}: ${resp.task_id}`);
         }
       } else {
         const payload: any = {