diff --git a/tools/link_detection/README_CN.md b/tools/link_detection/README_CN.md
new file mode 100644
index 0000000000000000000000000000000000000000..58ab4f8d238ee6be7c420320d17c41a82ab100f1
--- /dev/null
+++ b/tools/link_detection/README_CN.md
@@ -0,0 +1,31 @@
+# 链接检查工具
+
+## 简介
+
+此工具可以检查用户指定目录里所有文件的链接，将所有链接分为三类，并且将检查结果分别写入三个文件，如下所示：
+1. 响应的状态码不是200的链接，写入`400.txt`文件中。
+2. 脚本执行过程中请求出现异常的链接，写入`exception.txt`文件中。
+3. 对于安装包的链接，因为请求非常耗时，所以不发请求，直接写入`slow.txt`文件中。
+
+
+## 使用说明
+
+该工具所依赖的操作系统为Windows操作系统，执行环境为Python环境，具体使用步骤如下所示：
+
+1. 打开Git Bash，下载MindSpore Docs仓代码。
+   ```
+   git clone https://gitee.com/mindspore/docs.git
+   ```
+2. 进入`tools/link_detection`目录，安装执行所需的第三方库。
+   ```
+   cd tools/link_detection
+   pip install requests
+   ```
+3. 在`link_check`目录下执行如下命令，在输入需要检测目录的绝对路径后，开始进行检测，完成后会在当前目录下新建`404.txt`、`exception.txt`、`slow.txt`三个文件。
+   ```
+   python link_detection.py
+   ```
+   > 检测目录的绝对路径全使用英文，并且使用Linux的绝对路径方式，例如：`/d/master/docs`。 
+   
+
+
diff --git a/tools/link_detection/link_detection.py b/tools/link_detection/link_detection.py
new file mode 100644
index 0000000000000000000000000000000000000000..b2ff4f5cc344244b8cf2d2512dc6286e9881fea3
--- /dev/null
+++ b/tools/link_detection/link_detection.py
@@ -0,0 +1,101 @@
+import subprocess
+import re
+import requests
+import urllib3
+from concurrent.futures import ThreadPoolExecutor
+from threading import Lock
+
+def get_all_file(check_path):
+    '''
+    get all the files in the directory.
+    '''
+    cmd = 'find %s -type f' %check_path
+    res = subprocess.Popen(cmd, shell=True, stderr=subprocess.PIPE, stdout=subprocess.PIPE)
+    all_file_list = res.stdout.read().decode('utf-8').split('\n')
+    del all_file_list[-1]
+    return all_file_list
+
+def get_all_link(all_file_list):
+    '''
+    get all the links in all the files.
+    '''
+    re_rule = "(https:\/\/)([\w\-\.,@?^=%&:/~\+#]*[\w\-\@?^=%&/~\+#])?"
+    for i in all_file_list:
+        i = i.split('/', 1)[1].replace('/', ':/', 1)
+        try:
+            with open(i, 'r', encoding='utf-8') as f:
+                data = f.read()
+                link_list = []
+                urls = re.findall(re_rule, data, re.S)
+                if urls:
+                    for url in urls:
+                        link_list.append(url[0]+url[1])
+                if link_list:
+                    dic[i] = link_list
+        except Exception:
+            continue
+
+def get_status(addr):
+    '''
+    Request the link and write different results to different files.
+    '''
+    try:
+        link_path, link_addr, file_404, file_exception, mutexA, mutexB = addr[0], addr[1], addr[2], addr[3], addr[4], addr[5]
+        response = requests.get(link_addr, headers=headers, verify=False, timeout=5)
+        print(link_addr)
+        print(response.status_code)
+        if response.status_code != 200:
+            mutexA.acquire()
+            file_404.write('链接所在路径: %s' %link_path)
+            file_404.write('\n')
+            file_404.write('链接地址：%s' %link_addr)
+            file_404.write('\n')
+            file_404.write('链接的状态码：%s' %response.status_code)
+            file_404.write('\n\n\n\n\n')
+            mutexA.release()
+    except Exception :
+        print('exception!')
+        mutexB.acquire()
+        file_exception.write('链接所在路径: %s' %link_path)
+        file_exception.write('\n')
+        file_exception.write('链接地址：%s' %link_addr)
+        file_exception.write('\n\n\n\n\n')
+        mutexB.release()
+
+def multi_threading():
+    '''
+    open multithreading to finish tasks concurrently, do not send a request to the download link, write it directly.
+    '''
+    for i in dic:
+        link_list = list(set(dic[i]))
+        for j in link_list:
+            if j.endswith('.whl') or j.endswith('.gz'):
+                f3.write('链接所在路径: %s' %i)
+                f3.write('\n')
+                f3.write('链接地址：%s' %j)
+                f3.write('\n\n\n\n\n')
+                continue
+            pool.submit(get_status, (i, j, f1, f2, mutexA, mutexB))
+    pool.shutdown()
+    f1.close()
+    f2.close()
+    f3.close()
+
+def main():
+    all_file_list = get_all_file(check_path)
+    get_all_link(all_file_list)
+    multi_threading()
+
+if __name__ == '__main__':
+    check_path = input('请输入您要检测的绝对路径：').strip()
+    dic = {}
+    urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
+    headers = {
+        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.105 Safari/537.36'}
+    pool = ThreadPoolExecutor(500)
+    f1 = open('./404.txt', 'w', encoding='utf-8')
+    f2 = open('./exception.txt', 'w', encoding='utf-8')
+    f3 = open('./slow.txt', 'w', encoding='utf-8')
+    mutexA = Lock()
+    mutexB = Lock()
+    main()