diff --git a/src/en/guide/in_depth_guides/breakpoint_recovery.md b/src/en/guide/in_depth_guides/breakpoint_recovery.md
index f64b6fc..b94cc7f 100644
--- a/src/en/guide/in_depth_guides/breakpoint_recovery.md
+++ b/src/en/guide/in_depth_guides/breakpoint_recovery.md
@@ -29,7 +29,9 @@ When the program is interrupted or terminated, the file structure in the storage
       team.json          # Contains information such as team, environment, roles, actions, etc.
 ```
 
-Example of data summary corresponding to `team.json`.
+<details>
+
+<summary>Example of data summary corresponding to team.json </summary>
 
 ```json
 {
@@ -140,6 +142,8 @@ Example of data summary corresponding to `team.json`.
 }
 ```
 
+</details>
+
 ### Execution order during recovery
 
 Since MetaGPT is an asynchronous execution framework, there are several typical interception points and recovery sequences as follows.
diff --git a/src/en/guide/in_depth_guides/rag_module.md b/src/en/guide/in_depth_guides/rag_module.md
index aeddca0..612f3ef 100644
--- a/src/en/guide/in_depth_guides/rag_module.md
+++ b/src/en/guide/in_depth_guides/rag_module.md
@@ -10,7 +10,7 @@ This article focuses on the RAG functions provided by the current MetaGPT:
 4. Data update, addition of text and Python objects.
 5. Data storage and recovery, vectorization is not required each time.
 
-For more examples, please see [rag_pipeline](https://github.com/geekan/MetaGPT/blob/main/examples/rag_pipeline.py) and [rag_search](https://github.com/geekan/MetaGPT/blob/main/examples/rag_search.py)
+For more examples, please see [rag_pipeline](https://github.com/geekan/MetaGPT/blob/main/examples/rag/rag_pipeline.py) and [rag_search](https://github.com/geekan/MetaGPT/blob/main/examples/rag/rag_search.py)
 
 ## Prepare
 
diff --git a/src/zh/guide/in_depth_guides/breakpoint_recovery.md b/src/zh/guide/in_depth_guides/breakpoint_recovery.md
index 7a6245d..7cd33e5 100644
--- a/src/zh/guide/in_depth_guides/breakpoint_recovery.md
+++ b/src/zh/guide/in_depth_guides/breakpoint_recovery.md
@@ -29,7 +29,9 @@
       team.json          # 包含团队、环境、角色、动作等信息
 ```
 
-`team.json`对应内容的数据概要示例。
+<details>
+
+<summary>team.json 对应内容的数据概要示例。</summary>
 
 ```json
 {
@@ -140,6 +142,8 @@
 }
 ```
 
+</details>
+
 ### 恢复时的执行顺序
 
 由于MetaGPT是异步执行框架，对于下述几种典型的中断截点和恢复顺序。
diff --git a/src/zh/guide/in_depth_guides/rag_module.md b/src/zh/guide/in_depth_guides/rag_module.md
index 4fb476b..1f21b25 100644
--- a/src/zh/guide/in_depth_guides/rag_module.md
+++ b/src/zh/guide/in_depth_guides/rag_module.md
@@ -10,7 +10,7 @@ RAG（Retrieval-Augmented Generation）通过引用外部权威知识库来优
 4. 数据更新，增加文本与python对象
 5. 数据保存及恢复，不用每次都进行向量化
 
-更多的例子请查看 [rag_pipeline](https://github.com/geekan/MetaGPT/blob/main/examples/rag_pipeline.py) 和 [rag_search](https://github.com/geekan/MetaGPT/blob/main/examples/rag_search.py)
+更多的例子请查看 [rag_pipeline](https://github.com/geekan/MetaGPT/blob/main/examples/rag/rag_pipeline.py) 和 [rag_search](https://github.com/geekan/MetaGPT/blob/main/examples/rag/rag_search.py)
 
 ## 前置准备