diff --git a/dsRag/Config/Config.py b/dsRag/Config/Config.py
index 520bee08..db0bf35c 100644
--- a/dsRag/Config/Config.py
+++ b/dsRag/Config/Config.py
@@ -13,3 +13,11 @@ WORD2VEC_MODEL_PATH = r"D:\Tencent_AILab_ChineseEmbedding\Tencent_AILab_ChineseE
 # DeepSeek
 DEEPSEEK_API_KEY = 'sk-44ae895eeb614aa1a9c6460579e322f1'
 DEEPSEEK_URL = 'https://api.deepseek.com'
+
+
+# MYSQL配置信息
+MYSQL_HOST = "10.10.14.210"
+MYSQL_PORT = 22066
+MYSQL_USER = "root"
+MYSQL_PASSWORD = "DsideaL147258369"
+MYSQL_DB_NAME = "base_db"
\ No newline at end of file
diff --git a/dsRag/Config/__pycache__/Config.cpython-310.pyc b/dsRag/Config/__pycache__/Config.cpython-310.pyc
index 6451932c..a05d9cfa 100644
Binary files a/dsRag/Config/__pycache__/Config.cpython-310.pyc and b/dsRag/Config/__pycache__/Config.cpython-310.pyc differ
diff --git a/dsRag/Sql/t_ai_kb.sql b/dsRag/Sql/t_ai_kb.sql
new file mode 100644
index 00000000..d83cbfbe
--- /dev/null
+++ b/dsRag/Sql/t_ai_kb.sql
@@ -0,0 +1,61 @@
+/*
+ Navicat Premium Dump SQL
+
+ Source Server         : 10.10.14.210
+ Source Server Type    : MySQL
+ Source Server Version : 50742 (5.7.42-log)
+ Source Host           : 10.10.14.210:22066
+ Source Schema         : base_db
+
+ Target Server Type    : MySQL
+ Target Server Version : 50742 (5.7.42-log)
+ File Encoding         : 65001
+
+ Date: 24/06/2025 18:45:23
+*/
+
+SET NAMES utf8mb4;
+SET FOREIGN_KEY_CHECKS = 0;
+
+-- ----------------------------
+-- Table structure for t_ai_kb
+-- ----------------------------
+DROP TABLE IF EXISTS `t_ai_kb`;
+CREATE TABLE `t_ai_kb`  (
+  `id` int(11) NOT NULL AUTO_INCREMENT COMMENT '主键',
+  `kb_name` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NOT NULL COMMENT '知识库名称',
+  `short_name` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NOT NULL COMMENT '英文简称',
+  `create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
+  `is_delete` int(11) NOT NULL DEFAULT 0 COMMENT '是否删除',
+  PRIMARY KEY (`id`) USING BTREE,
+  UNIQUE INDEX `short_name`(`short_name`) USING BTREE,
+  INDEX `is_delete`(`is_delete`) USING BTREE
+) ENGINE = InnoDB AUTO_INCREMENT = 1 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_general_ci COMMENT = 'AI知识库' ROW_FORMAT = Dynamic;
+
+-- ----------------------------
+-- Records of t_ai_kb
+-- ----------------------------
+
+-- ----------------------------
+-- Table structure for t_ai_kb_files
+-- ----------------------------
+DROP TABLE IF EXISTS `t_ai_kb_files`;
+CREATE TABLE `t_ai_kb_files`  (
+  `id` int(11) NOT NULL AUTO_INCREMENT COMMENT '主键',
+  `file_name` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NOT NULL COMMENT '文件名称',
+  `ext_name` varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NOT NULL COMMENT '文件扩展名',
+  `create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
+  `kb_id` int(11) NOT NULL COMMENT '隶属知识库ID',
+  `is_delete` int(11) NOT NULL DEFAULT 0 COMMENT '是否删除',
+  `state` int(11) NOT NULL DEFAULT 0 COMMENT '0:上传后未处理，1：上传后已处理,2:处理失败',
+  PRIMARY KEY (`id`) USING BTREE,
+  INDEX `kb_id`(`kb_id`) USING BTREE,
+  INDEX `is_delete`(`is_delete`, `state`) USING BTREE,
+  CONSTRAINT `t_ai_kb_files_ibfk_1` FOREIGN KEY (`kb_id`) REFERENCES `t_ai_kb` (`id`) ON DELETE CASCADE ON UPDATE RESTRICT
+) ENGINE = InnoDB AUTO_INCREMENT = 1 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_general_ci COMMENT = 'AI知识库上传的文件' ROW_FORMAT = Dynamic;
+
+-- ----------------------------
+-- Records of t_ai_kb_files
+-- ----------------------------
+
+SET FOREIGN_KEY_CHECKS = 1;
diff --git a/dsRag/Start.py b/dsRag/Start.py
new file mode 100644
index 00000000..1c452d94
--- /dev/null
+++ b/dsRag/Start.py
@@ -0,0 +1,137 @@
+"""
+pip install fastapi uvicorn aiomysql
+"""
+import uvicorn
+from fastapi import FastAPI, HTTPException
+from pydantic import BaseModel
+
+from Util.MySQLUtil import *
+
+"""
+API文档访问： http://localhost:8000/docs
+该实现包含以下功能：
+
+- 知识库(t_ai_kb)的增删改查接口
+- 知识库文件(t_ai_kb_files)的增删改查接口
+- 使用MySQLUtil.py中的连接池管理
+- 自动生成的Swagger文档
+"""
+
+app = FastAPI()
+
+# 知识库模型
+class KbModel(BaseModel):
+    kb_name: str
+    short_name: str
+    is_delete: Optional[int] = 0
+
+# 知识库文件模型
+class KbFileModel(BaseModel):
+    file_name: str
+    ext_name: str
+    kb_id: int
+    is_delete: Optional[int] = 0
+    state: Optional[int] = 0
+
+@app.on_event("startup")
+async def startup_event():
+    app.state.mysql_pool = await init_mysql_pool()
+
+@app.on_event("shutdown")
+async def shutdown_event():
+    app.state.mysql_pool.close()
+    await app.state.mysql_pool.wait_closed()
+
+# 知识库CRUD接口
+@app.post("/kb")
+async def create_kb(kb: KbModel):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor() as cur:
+            await cur.execute(
+                """INSERT INTO t_ai_kb (kb_name, short_name, is_delete) 
+                VALUES (%s, %s, %s)""",
+                (kb.kb_name, kb.short_name, kb.is_delete)
+            )
+            await conn.commit()
+            return {"id": cur.lastrowid}
+
+@app.get("/kb/{kb_id}")
+async def read_kb(kb_id: int):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor(DictCursor) as cur:
+            await cur.execute("SELECT * FROM t_ai_kb WHERE id = %s", (kb_id,))
+            result = await cur.fetchone()
+            if not result:
+                raise HTTPException(status_code=404, detail="Knowledge base not found")
+            return result
+
+@app.put("/kb/{kb_id}")
+async def update_kb(kb_id: int, kb: KbModel):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor() as cur:
+            await cur.execute(
+                """UPDATE t_ai_kb 
+                SET kb_name = %s, short_name = %s, is_delete = %s 
+                WHERE id = %s""",
+                (kb.kb_name, kb.short_name, kb.is_delete, kb_id)
+            )
+            await conn.commit()
+            return {"message": "Knowledge base updated"}
+
+@app.delete("/kb/{kb_id}")
+async def delete_kb(kb_id: int):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor() as cur:
+            await cur.execute("DELETE FROM t_ai_kb WHERE id = %s", (kb_id,))
+            await conn.commit()
+            return {"message": "Knowledge base deleted"}
+
+# 知识库文件CRUD接口
+@app.post("/kb_files")
+async def create_kb_file(file: KbFileModel):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor() as cur:
+            await cur.execute(
+                """INSERT INTO t_ai_kb_files 
+                (file_name, ext_name, kb_id, is_delete, state) 
+                VALUES (%s, %s, %s, %s, %s)""",
+                (file.file_name, file.ext_name, file.kb_id, file.is_delete, file.state)
+            )
+            await conn.commit()
+            return {"id": cur.lastrowid}
+
+@app.get("/kb_files/{file_id}")
+async def read_kb_file(file_id: int):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor(DictCursor) as cur:
+            await cur.execute("SELECT * FROM t_ai_kb_files WHERE id = %s", (file_id,))
+            result = await cur.fetchone()
+            if not result:
+                raise HTTPException(status_code=404, detail="File not found")
+            return result
+
+@app.put("/kb_files/{file_id}")
+async def update_kb_file(file_id: int, file: KbFileModel):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor() as cur:
+            await cur.execute(
+                """UPDATE t_ai_kb_files 
+                SET file_name = %s, ext_name = %s, kb_id = %s, 
+                    is_delete = %s, state = %s 
+                WHERE id = %s""",
+                (file.file_name, file.ext_name, file.kb_id, 
+                 file.is_delete, file.state, file_id)
+            )
+            await conn.commit()
+            return {"message": "File updated"}
+
+@app.delete("/kb_files/{file_id}")
+async def delete_kb_file(file_id: int):
+    async with app.state.mysql_pool.acquire() as conn:
+        async with conn.cursor() as cur:
+            await cur.execute("DELETE FROM t_ai_kb_files WHERE id = %s", (file_id,))
+            await conn.commit()
+            return {"message": "File deleted"}
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=8000)
\ No newline at end of file
diff --git a/dsRag/Util/MySQLUtil.py b/dsRag/Util/MySQLUtil.py
new file mode 100644
index 00000000..0302e558
--- /dev/null
+++ b/dsRag/Util/MySQLUtil.py
@@ -0,0 +1,281 @@
+"""
+pip install aiomysql
+"""
+import logging
+from typing import Optional, Dict, List
+
+from aiomysql import create_pool
+from Config.Config import *
+# 配置日志
+logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")
+logger = logging.getLogger(__name__)
+
+# MySQL 配置
+MYSQL_CONFIG = {
+    "host": MYSQL_HOST,
+    "port": MYSQL_PORT,
+    "user": MYSQL_USER,
+    "password": MYSQL_PASSWORD,
+    "db": MYSQL_DB_NAME,
+    "minsize": 1,
+    "maxsize": 20,
+}
+
+
+# 初始化 MySQL 连接池
+async def init_mysql_pool():
+    return await create_pool(**MYSQL_CONFIG)
+
+
+# 保存聊天记录到 MySQL
+async def save_chat_to_mysql(mysql_pool, person_id, prompt, result, audio_url, duration, input_type=1, output_type=1,
+                             input_image_type=0, image_width=0, image_height=0):
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cur:
+            await cur.execute(
+                "INSERT INTO t_chat_log (person_id, user_input, model_response,audio_url,duration,input_type,output_type,input_image_type,image_width,image_height,create_time) VALUES (%s, %s, %s, %s, %s, %s, %s,%s,%s,%s,NOW())",
+                (person_id, prompt, result, audio_url, duration, input_type, output_type, input_image_type, image_width,
+                 image_height)
+            )
+            await conn.commit()
+
+
+# 清空表
+async def truncate_chat_log(mysql_pool):
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cur:
+            await cur.execute("TRUNCATE TABLE t_chat_log")
+            await conn.commit()
+            logger.info("表 t_chat_log 已清空。")
+
+
+from aiomysql import DictCursor
+
+
+# 分页查询聊天记录
+async def get_chat_log_by_session(mysql_pool, person_id, page=1, page_size=10):
+    """
+    根据 person_id 查询聊天记录，并按 id 降序分页
+    :param mysql_pool: MySQL 连接池
+    :param person_id: 用户会话 ID
+    :param page: 当前页码（默认值为 1，但会动态计算为最后一页）
+    :param page_size: 每页记录数
+    :return: 分页数据
+    """
+    if not mysql_pool:
+        raise ValueError("MySQL 连接池未初始化")
+
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor(DictCursor) as cur:  # 使用 DictCursor
+            # 查询总记录数
+            await cur.execute(
+                "SELECT COUNT(*) FROM t_chat_log WHERE person_id = %s",
+                (person_id,)
+            )
+            total = (await cur.fetchone())['COUNT(*)']
+
+            # 计算总页数
+            total_pages = (total + page_size - 1) // page_size
+
+            # 计算偏移量
+            offset = (page - 1) * page_size
+
+            # 查询分页数据，按 id 降序排列
+            await cur.execute(
+                "SELECT id, person_id, user_input, model_response, audio_url, duration,input_type,output_type,input_image_type,image_width,image_height, create_time "
+                "FROM t_chat_log WHERE person_id = %s ORDER BY id DESC LIMIT %s OFFSET %s",
+                (person_id, page_size, offset)
+            )
+            records = await cur.fetchall()
+
+            # 将查询结果反转，确保最新消息显示在最后
+            if page==1 and records:
+                records.reverse()
+
+            # 将查询结果转换为字典列表
+            result = [
+                {
+                    "id": record['id'],
+                    "person_id": record['person_id'],
+                    "user_input": record['user_input'],
+                    "model_response": record['model_response'],
+                    "audio_url": record['audio_url'],
+                    "duration": record['duration'],
+                    "input_type": record['input_type'],
+                    "output_type": record['output_type'],
+                    "image_width": record['image_width'],
+                    "image_height": record['image_height'],
+                    "input_image_type": record['input_image_type'],
+                    "create_time": record['create_time'].strftime("%Y-%m-%d %H:%M:%S")
+                }
+                for record in records
+            ]
+
+            return {
+                "data": result,  # 按 id 升序排列的数据
+                "total": total,
+                "page": page,
+                "page_size": page_size,
+                "total_pages": total_pages
+            }
+
+
+# 获取指定会话的最后一条记录的 id
+async def get_last_chat_log_id(mysql_pool, person_id):
+    """
+    获取指定会话的最后一条记录的 id
+    :param mysql_pool: MySQL 连接池
+    :param session_id: 用户会话 ID
+    :return: 最后一条记录的 id，如果未找到则返回 None
+    """
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cur:
+            await cur.execute(
+                "SELECT id FROM t_chat_log WHERE person_id = %s ORDER BY id DESC LIMIT 1",
+                (person_id,)
+            )
+            result = await cur.fetchone()
+            return result[0] if result else None
+
+
+# 更新为危险的记录
+async def update_risk(mysql_pool, person_id, risk_memo):
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cur:
+            # 1. 获取此人员的最后一条记录 id
+            last_id = await get_last_chat_log_id(mysql_pool, person_id)
+
+            if last_id:
+                # 2. 更新 risk_flag 和 risk_memo
+                await cur.execute(
+                    "UPDATE t_chat_log SET risk_flag = 1, risk_memo = %s WHERE id = %s",
+                    (risk_memo.replace('\n', '').replace("NO", ""), last_id)
+                )
+                await conn.commit()
+                logger.info(f"已更新 person_id={person_id} 的最后一条记录 (id={last_id}) 的 risk_flag 和 risk_memo。")
+            else:
+                logger.warning(f"未找到 person_id={person_id} 的记录。")
+
+
+# 查询用户信息
+async def get_user_by_login_name(mysql_pool, login_name: str) -> Optional[Dict]:
+    """
+    根据用户名查询用户信息
+    :param pool: MySQL 连接池
+    :param login_name: 用户名
+    :return: 用户信息（字典形式）
+    """
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cursor:
+            sql = "SELECT * FROM t_base_person WHERE login_name = %s"
+            await cursor.execute(sql, (login_name,))
+            row = await cursor.fetchone()
+            if not row:
+                return None
+
+            # 将元组转换为字典
+            columns = [column[0] for column in cursor.description]
+            return dict(zip(columns, row))
+
+
+# 显示统计分析页面
+async def get_chat_logs_summary(mysql_pool, risk_flag: int, offset: int, page_size: int) -> (List[Dict], int):
+    """
+    获取聊天记录的统计分析结果
+    :param mysql_pool: MySQL 连接池
+    :param risk_flag: 风险标志
+    :param offset: 偏移量
+    :param page_size: 每页记录数
+    :return: 日志列表和总记录数
+    """
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cursor:
+            # 查询符合条件的记录
+            sql = """
+                SELECT tbp.*, COUNT(*) AS cnt 
+                FROM t_chat_log AS tcl 
+                INNER JOIN t_base_person AS tbp ON tcl.person_id = tbp.person_id
+                WHERE tcl.risk_flag = %s 
+                GROUP BY tcl.person_id 
+                ORDER BY COUNT(*) DESC
+                LIMIT %s OFFSET %s
+            """
+            await cursor.execute(sql, (risk_flag, page_size, offset))
+            rows = await cursor.fetchall()
+
+            # 获取列名
+            columns = [column[0] for column in cursor.description]
+
+            # 查询总记录数
+            count_sql = """
+                SELECT COUNT(DISTINCT tcl.person_id) 
+                FROM t_chat_log AS tcl 
+                INNER JOIN t_base_person AS tbp ON tcl.person_id = tbp.person_id 
+                WHERE tcl.risk_flag = %s
+            """
+            await cursor.execute(count_sql, (risk_flag,))
+            total = (await cursor.fetchone())[0]
+
+            # 将元组转换为字典
+            logs = [dict(zip(columns, row)) for row in rows] if rows else []
+
+            return logs, total
+
+
+async def get_chat_logs_by_risk_flag(mysql_pool, risk_flag: int, person_id: str, offset: int, page_size: int) -> (
+        List[Dict], int):
+    """
+    根据风险标志查询聊天记录
+    :param mysql_pool: MySQL 连接池
+    :param risk_flag: 风险标志
+    :param offset: 分页偏移量
+    :param page_size: 每页记录数
+    :return: 聊天记录列表和总记录数
+    """
+    async with mysql_pool.acquire() as conn:
+        await conn.ping()  # 重置连接
+        async with conn.cursor() as cursor:
+            # 查询符合条件的记录
+            sql = """
+                SELECT  tcl.id, tcl.user_input, tcl.model_response, tcl.audio_url, tcl.duration, 
+                    tcl.create_time, tcl.risk_flag, tcl.risk_memo, tcl.risk_result, 
+                    tbp.person_id, tbp.login_name, tbp.person_name 
+                FROM t_chat_log AS tcl 
+                INNER JOIN t_base_person AS tbp ON tcl.person_id = tbp.person_id 
+                WHERE tcl.risk_flag = %s and tcl.person_id=%s ORDER BY TCL.ID DESC 
+                LIMIT %s OFFSET %s
+            """
+            await cursor.execute(sql, (risk_flag, person_id, page_size, offset))
+            rows = await cursor.fetchall()
+
+            # 在 count_sql 执行前获取列名
+            columns = [column[0] for column in cursor.description]
+
+            # 查询总记录数
+            count_sql = """
+                SELECT COUNT(*) 
+                FROM t_chat_log AS tcl 
+                INNER JOIN t_base_person AS tbp ON tcl.person_id = tbp.person_id 
+                WHERE tcl.risk_flag = %s and tcl.person_id=%s
+            """
+            await cursor.execute(count_sql, (risk_flag, person_id))
+            total = (await cursor.fetchone())[0]
+
+            # 将元组转换为字典，并格式化 create_time
+            if rows:
+                logs = []
+                for row in rows:
+                    log = dict(zip(columns, row))
+                    # 格式化 create_time
+                    if log["create_time"]:
+                        log["create_time"] = log["create_time"].strftime("%Y-%m-%d %H:%M:%S")
+                    logs.append(log)
+                return logs, total
+            return [], 0
diff --git a/dsRag/Util/__pycache__/MySQLUtil.cpython-310.pyc b/dsRag/Util/__pycache__/MySQLUtil.cpython-310.pyc
new file mode 100644
index 00000000..c310e899
Binary files /dev/null and b/dsRag/Util/__pycache__/MySQLUtil.cpython-310.pyc differ