feat: 入库改为独立页面，更新 API 说明文档与 Agent skill

- 新增 Ingest.vue 独立入库页（文本/文件/批量 + 任务进度列表），移除入库 Drawer
- 导航栏去掉「入库进度」，新增「文档入库」入口；知识库页「入库」跳转独立页
- API 说明页接口清单与后端对齐（补齐用户管理/上传批量/任务重试删除/reingest 等，修正鉴权标注）
- 更新 qmdsearch-agent skill 与 api-examples（鉴权说明、会话 token、新端点示例）

skills/qmdsearch-agent/references/api-examples.md

@@ -36,11 +36,32 @@ curl -s -X POST {BASE}/api/v1/documents/upload \
 curl -s {BASE}/api/v1/documents/tasks/<task_id>
 ```
 
-### 检索
+### 批量文件上传（202，逐文件建任务）
+
+```bash
+curl -s -X POST {BASE}/api/v1/documents/upload-batch \
+  -H "Authorization: Bearer <token>" \
+  -F "files=@a.pdf" -F "files=@b.md"
+```
+
+### 列出入库任务 / 重试 / 删除任务
+
+```bash
+curl -s -H "Authorization: Bearer <token>" "{BASE}/api/v1/documents/tasks?limit=20"
+curl -s -X POST -H "Authorization: Bearer <token>" {BASE}/api/v1/documents/tasks/<task_id>/retry
+curl -s -X DELETE -H "Authorization: Bearer <token>" {BASE}/api/v1/documents/tasks/<task_id>
+```
+
+### 重新摘要入库（读原文件 → 删旧 → 重跑）
+
+```bash
+curl -s -X POST -H "Authorization: Bearer <token>" {BASE}/api/v1/documents/<doc_id>/reingest
+```
+
+### 检索（免登录）
 
 ```bash
 curl -s -X POST {BASE}/api/v1/search \
-  -H "Authorization: Bearer <token>" \
   -H "Content-Type: application/json" \
   -d '{"query":"如何配置 Redis 缓存","top_k":5,"summarize":false}'
 ```
@@ -120,6 +141,13 @@ class QMDSearchClient:
                              headers=self._headers(), files=files, data=data, timeout=60)
         return self._ok(resp)["task_id"]
 
+    def upload_batch(self, paths: list[str]) -> dict:
+        """批量上传多个文件：返回 {tasks:[{filename,task_id}], failed:[{filename,error}]}"""
+        files = [("files", open(p, "rb")) for p in paths]
+        resp = requests.post(f"{self.base}/api/v1/documents/upload-batch",
+                             headers=self._headers(), files=files, timeout=60)
+        return self._ok(resp)
+
     # ---- 轮询 ----
     def wait_task(self, task_id: str) -> dict:
         deadline = time.time() + self.poll_timeout
@@ -165,6 +193,6 @@ if __name__ == "__main__":
 
 ## 三、响应结构速记
 
-- 文本 / 文件入库 `202` → `data = {task_id, status:"pending"[, saved_path]}`
+- 文本 / 文件 / 批量入库 `202` → `data = {task_id, status:"pending"[, saved_path]}`；批量返回 `{tasks, failed}`
 - 任务 `done` → `data.result = {document_id, summary:{l1_summary,l2_outline,l3_content_outline,level}, category, chunks_count, tags, category_confidence, deduplicated}`
 - 检索 → `data = {query, hits:[{text,doc_id,title,section_path,score,doc_summary}], routed_categories, fallback, extracted_info, summary?}`