langchain-dev-utils 1.2.5__tar.gz → 1.2.7__tar.gz

{langchain_dev_utils-1.2.5 → langchain_dev_utils-1.2.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: langchain-dev-utils
-Version: 1.2.5
+Version: 1.2.7
 Summary: A practical utility library for LangChain and LangGraph development
 Project-URL: Source Code, https://github.com/TBice123123/langchain-dev-utils
 Project-URL: repository, https://github.com/TBice123123/langchain-dev-utils
@@ -11,6 +11,7 @@ Requires-Python: >=3.11
 Requires-Dist: langchain>=1.1.0
 Requires-Dist: langgraph>=1.0.0
 Provides-Extra: standard
+Requires-Dist: json-repair>=0.53.1; extra == 'standard'
 Requires-Dist: langchain-openai; extra == 'standard'
 Description-Content-Type: text/markdown
 
@@ -57,24 +58,6 @@ Mainly consists of the following two functions:
 - `register_model_provider`: Register a chat model provider
 - `load_chat_model`: Load a chat model
 
-**`register_model_provider` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `provider_name` | str | Yes | - | The name of the model provider, used as an identifier for loading models later. |
-| `chat_model` | ChatModel \| str | Yes | - | The chat model, which can be either a `ChatModel` instance or a string (currently only `"openai-compatible"` is supported). |
-| `base_url` | str | No | - | The API endpoint URL of the model provider (applicable to both `chat_model` types, but primarily used when `chat_model` is a string with value `"openai-compatible"`). |
-| `model_profiles` | dict | No | - | Declares the capabilities and parameters supported by each model provided by this provider. The configuration corresponding to the `model_name` will be loaded and assigned to `model.profile` (e.g., fields such as `max_input_tokens`, `tool_calling` etc.). |
-| `compatibility_options` | dict | No | - | Compatibility options for the model provider (only effective when `chat_model` is a string with value `"openai-compatible"`). Used to declare support for OpenAI-compatible features (e.g., `tool_choice` strategies, JSON mode, etc.) to ensure correct functional adaptation. |
-
-**`load_chat_model` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `model` | str | Yes | - | Chat model name |
-| `model_provider` | str | No | - | Chat model provider name |
-| `kwargs` | dict | No | - | Additional parameters passed to the chat model class, e.g., temperature, top_p, etc. |
-
 Example for integrating a qwen3-4b model deployed using `vllm`:
 
 ```python
@@ -102,22 +85,6 @@ Mainly consists of the following two functions:
 - `register_embeddings_provider`: Register an embedding model provider
 - `load_embeddings`: Load an embedding model
 
-**`register_embeddings_provider` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `provider_name` | str | Yes | - | Embedding model provider name, used as an identifier for subsequent model loading |
-| `embeddings_model` | Embeddings \| str | Yes | - | Embedding model, can be Embeddings or a string (currently supports "openai-compatible") |
-| `base_url` | str | No | - | The API address of the Embedding model provider (valid for both types of `embeddings_model`, but mainly used when `embeddings_model` is a string and is "openai-compatible") |
-
-**`load_embeddings` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `model` | str | Yes | - | Embedding model name |
-| `provider` | str | No | - | Embedding model provider name |
-| `kwargs` | dict | No | - | Other additional parameters |
-
 Example for integrating a qwen3-embedding-4b model deployed using `vllm`:
 
 ```python
@@ -150,13 +117,9 @@ Includes the following features:
 
 For stream responses obtained using `stream()` and `astream()`, you can use `merge_ai_message_chunk` to merge them into a final AIMessage.
 
-**`merge_ai_message_chunk` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `chunks` | List[AIMessageChunk] | Yes | - | List of AIMessageChunk objects |
-
 ```python
+from langchain_dev_utils.message_convert import merge_ai_message_chunk
+
 chunks = list(model.stream("Hello"))
 merged = merge_ai_message_chunk(chunks)
 ```
@@ -165,15 +128,8 @@ merged = merge_ai_message_chunk(chunks)
 
 For a list, you can use `format_sequence` to format it.
 
-**`format_sequence` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `inputs` | List | Yes | - | A list containing any of the following types: langchain_core.messages, langchain_core.documents.Document, str |
-| `separator` | str | No | "-" | String used to join the content |
-| `with_num` | bool | No | False | If True, add a numeric prefix to each item (e.g., "1. Hello") |
-
 ```python
+from langchain_dev_utils.message_convert import format_sequence
 text = format_sequence([
     "str1",
     "str2",
@@ -194,19 +150,6 @@ Includes the following features:
 
 `has_tool_calling` and `parse_tool_calling` are used to check and parse tool calls.
 
-**`has_tool_calling` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `message` | AIMessage | Yes | - | AIMessage object |
-
-**`parse_tool_calling` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `message` | AIMessage | Yes | - | AIMessage object |
-| `first_tool_call_only` | bool | No | False | Whether to only check the first tool call |
-
 ```python
 import datetime
 from langchain_core.tools import tool
@@ -234,7 +177,7 @@ if has_tool_calling(response):
 Both can accept a `handler` parameter for custom breakpoint return and response handling logic.
 
 ```python
-from langchain_dev_utils import human_in_the_loop
+from langchain_dev_utils.tool_calling import human_in_the_loop
 from langchain_core.tools import tool
 import datetime
 
@@ -256,14 +199,7 @@ Includes the following features:
 
 #### 4.1 Agent Factory Functions
 
-In LangChain v1, the officially provided `create_agent` function can be used to create a single agent, where the model parameter supports passing a BaseChatModel instance or a specific string (when passing a string, it is limited to the models supported by `init_chat_model`). To extend the flexibility of specifying models via strings, this library provides a functionally identical `create_agent` function, allowing you to directly use models supported by `load_chat_model` (requires prior registration).
-
-**`create_agent` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `model` | str \| BaseChatModel | Yes | - | Model name or model instance. Can be a string identifier for a model registered with `register_model_provider` or a BaseChatModel instance. |
-| Other parameters | Various | No | - | All other parameters are the same as in `langchain.agents.create_agent` |
+In LangChain v1, the official `create_agent` function can be used to create a single agent; its `model` parameter accepts either a BaseChatModel instance or a specific string (when a string is provided, only models supported by `init_chat_model` are allowed). To extend the flexibility of specifying models via string, this library provides an equivalent `create_agent` function that lets you designate any model supported by `load_chat_model` (registration required beforehand).
 
 Usage example:
 
@@ -278,24 +214,26 @@ print(response)
 
 #### 4.2 Middleware
 
-Provides some commonly used middleware components. Below are examples of `SummarizationMiddleware` and `PlanMiddleware`.
+Provides some commonly used middleware components. Below, we illustrate with `ToolCallRepairMiddleware` and `PlanMiddleware`.
 
-`SummarizationMiddleware` is used for agent summarization.
+`ToolCallRepairMiddleware` is used to repair `invalid_tool_calls` generated by large language models.
 
 `PlanMiddleware` is used for agent planning.
 
 ```python
 from langchain_dev_utils.agents.middleware import (
-    SummarizationMiddleware,
+    ToolcallRepairMiddleware,
     PlanMiddleware,
 )
 
-agent=create_agent(
+agent = create_agent(
     "vllm:qwen3-4b",
     name="plan-agent",
-    middleware=[PlanMiddleware(), SummarizationMiddleware(model="vllm:qwen3-4b")]
+    middleware=[ToolCallRepairMiddleware(), PlanMiddleware(
+        use_read_plan_tool=False
+    )]
 )
-response = agent.invoke({"messages": [{"role": "user", "content": "Give me a travel plan to New York"}]}))
+response = agent.invoke({"messages": [{"role": "user", "content": "Give me a travel plan to New York"}]})
 print(response)
 ```
 
@@ -303,29 +241,14 @@ print(response)
 
 ### 5. **State Graph Orchestration**
 
-Includes the following features:
+Includes the following capabilities:
 
 - Sequential graph orchestration
 - Parallel graph orchestration
 
 #### 5.1 Sequential Graph Orchestration
 
-Sequential graph orchestration:
-Uses `create_sequential_pipeline`, supported parameters:
-
-**`create_sequential_pipeline` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `sub_graphs` | List[StateGraph] | Yes | - | List of state graphs to combine (must be StateGraph instances) |
-| `state_schema` | type | Yes | - | State Schema for the final generated graph |
-| `graph_name` | str | No | - | Name of the final generated graph |
-| `context_schema` | type | No | - | Context Schema for the final generated graph |
-| `input_schema` | type | No | - | Input Schema for the final generated graph |
-| `output_schema` | type | No | - | Output Schema for the final generated graph |
-| `checkpoint` | BaseCheckpointSaver | No | - | LangGraph persistence Checkpoint |
-| `store` | BaseStore | No | - | LangGraph persistence Store |
-| `cache` | BaseCache | No | - | LangGraph Cache |
+Use `create_sequential_pipeline` to orchestrate multiple subgraphs in sequential order:
 
 ```python
 from langchain.agents import AgentState
@@ -340,25 +263,25 @@ register_model_provider(
     base_url="http://localhost:8000/v1",
 )
 
-# Build sequential pipeline (all sub-graphs execute sequentially)
+# Build a sequential pipeline (all subgraphs executed in order)
 graph = create_sequential_pipeline(
     sub_graphs=[
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_time],
-            system_prompt="You are a time query assistant, can only answer the current time. If the question is unrelated to time, please directly answer that you cannot answer.",
+            system_prompt="You are a time-query assistant. You can only answer questions about the current time. If the question is unrelated to time, respond with 'I cannot answer that.'",
             name="time_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_weather],
-            system_prompt="You are a weather query assistant, can only answer the current weather. If the question is unrelated to weather, please directly answer that you cannot answer.",
+            system_prompt="You are a weather-query assistant. You can only answer questions about the current weather. If the question is unrelated to weather, respond with 'I cannot answer that.'",
             name="weather_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_user],
-            system_prompt="You are a user query assistant, can only answer the current user. If the question is unrelated to user, please directly answer that you cannot answer.",
+            system_prompt="You are a user-query assistant. You can only answer questions about the current user. If the question is unrelated to the user, respond with 'I cannot answer that.'",
             name="user_agent",
         ),
     ],
@@ -371,51 +294,36 @@ print(response)
 
 #### 5.2 Parallel Graph Orchestration
 
-Parallel graph orchestration:
-Uses `create_parallel_pipeline`, supported parameters:
-
-**`create_parallel_pipeline` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `sub_graphs` | List[StateGraph] | Yes | - | List of state graphs to combine |
-| `state_schema` | type | Yes | - | State Schema for the final generated graph |
-| `branches_fn` | Callable | Yes | - | Parallel branch function, returns a list of Send objects to control parallel execution |
-| `graph_name` | str | No | - | Name of the final generated graph |
-| `context_schema` | type | No | - | Context Schema for the final generated graph |
-| `input_schema` | type | No | - | Input Schema for the final generated graph |
-| `output_schema` | type | No | - | Output Schema for the final generated graph |
-| `checkpoint` | BaseCheckpointSaver | No | - | LangGraph persistence Checkpoint |
-| `store` | BaseStore | No | - | LangGraph persistence Store |
-| `cache` | BaseCache | No | - | LangGraph Cache |
+Use `create_parallel_pipeline` to orchestrate multiple subgraphs in parallel:
 
 ```python
 from langchain_dev_utils.pipeline import create_parallel_pipeline
 
-# Build parallel pipeline (all sub-graphs execute in parallel)
+# Build a parallel pipeline (all subgraphs executed concurrently)
 graph = create_parallel_pipeline(
     sub_graphs=[
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_time],
-            system_prompt="You are a time query assistant, can only answer the current time. If the question is unrelated to time, please directly answer that you cannot answer.",
+            system_prompt="You are a time-query assistant. You can only answer questions about the current time. If the question is unrelated to time, respond with 'I cannot answer that.'",
             name="time_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_weather],
-            system_prompt="You are a weather query assistant, can only answer the current weather. If the question is unrelated to weather, please directly answer that you cannot answer.",
+            system_prompt="You are a weather-query assistant. You can only answer questions about the current weather. If the question is unrelated to weather, respond with 'I cannot answer that.'",
             name="weather_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_user],
-            system_prompt="You are a user query assistant, can only answer the current user. If the question is unrelated to user, please directly answer that you cannot answer.",
+            system_prompt="You are a user-query assistant. You can only answer questions about the current user. If the question is unrelated to the user, respond with 'I cannot answer that.'",
             name="user_agent",
         ),
     ],
     state_schema=AgentState,
 )
+
 response = graph.invoke({"messages": [HumanMessage("Hello")]})
 print(response)
 ```
@@ -426,4 +334,4 @@ print(response)
 
 - [GitHub Repository](https://github.com/TBice123123/langchain-dev-utils) — Browse source code, submit Pull Requests
 - [Issue Tracker](https://github.com/TBice123123/langchain-dev-utils/issues) — Report bugs or suggest improvements
-- We welcome contributions in all forms — whether code, documentation, or usage examples. Let's build a more powerful and practical LangChain development ecosystem together!
+- We welcome contributions in all forms — whether code, documentation, or usage examples. Let's build a more powerful and practical LangChain development ecosystem together!

{langchain_dev_utils-1.2.5 → langchain_dev_utils-1.2.7}/README.md

@@ -41,24 +41,6 @@ Mainly consists of the following two functions:
 - `register_model_provider`: Register a chat model provider
 - `load_chat_model`: Load a chat model
 
-**`register_model_provider` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `provider_name` | str | Yes | - | The name of the model provider, used as an identifier for loading models later. |
-| `chat_model` | ChatModel \| str | Yes | - | The chat model, which can be either a `ChatModel` instance or a string (currently only `"openai-compatible"` is supported). |
-| `base_url` | str | No | - | The API endpoint URL of the model provider (applicable to both `chat_model` types, but primarily used when `chat_model` is a string with value `"openai-compatible"`). |
-| `model_profiles` | dict | No | - | Declares the capabilities and parameters supported by each model provided by this provider. The configuration corresponding to the `model_name` will be loaded and assigned to `model.profile` (e.g., fields such as `max_input_tokens`, `tool_calling` etc.). |
-| `compatibility_options` | dict | No | - | Compatibility options for the model provider (only effective when `chat_model` is a string with value `"openai-compatible"`). Used to declare support for OpenAI-compatible features (e.g., `tool_choice` strategies, JSON mode, etc.) to ensure correct functional adaptation. |
-
-**`load_chat_model` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `model` | str | Yes | - | Chat model name |
-| `model_provider` | str | No | - | Chat model provider name |
-| `kwargs` | dict | No | - | Additional parameters passed to the chat model class, e.g., temperature, top_p, etc. |
-
 Example for integrating a qwen3-4b model deployed using `vllm`:
 
 ```python
@@ -86,22 +68,6 @@ Mainly consists of the following two functions:
 - `register_embeddings_provider`: Register an embedding model provider
 - `load_embeddings`: Load an embedding model
 
-**`register_embeddings_provider` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `provider_name` | str | Yes | - | Embedding model provider name, used as an identifier for subsequent model loading |
-| `embeddings_model` | Embeddings \| str | Yes | - | Embedding model, can be Embeddings or a string (currently supports "openai-compatible") |
-| `base_url` | str | No | - | The API address of the Embedding model provider (valid for both types of `embeddings_model`, but mainly used when `embeddings_model` is a string and is "openai-compatible") |
-
-**`load_embeddings` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `model` | str | Yes | - | Embedding model name |
-| `provider` | str | No | - | Embedding model provider name |
-| `kwargs` | dict | No | - | Other additional parameters |
-
 Example for integrating a qwen3-embedding-4b model deployed using `vllm`:
 
 ```python
@@ -134,13 +100,9 @@ Includes the following features:
 
 For stream responses obtained using `stream()` and `astream()`, you can use `merge_ai_message_chunk` to merge them into a final AIMessage.
 
-**`merge_ai_message_chunk` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `chunks` | List[AIMessageChunk] | Yes | - | List of AIMessageChunk objects |
-
 ```python
+from langchain_dev_utils.message_convert import merge_ai_message_chunk
+
 chunks = list(model.stream("Hello"))
 merged = merge_ai_message_chunk(chunks)
 ```
@@ -149,15 +111,8 @@ merged = merge_ai_message_chunk(chunks)
 
 For a list, you can use `format_sequence` to format it.
 
-**`format_sequence` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `inputs` | List | Yes | - | A list containing any of the following types: langchain_core.messages, langchain_core.documents.Document, str |
-| `separator` | str | No | "-" | String used to join the content |
-| `with_num` | bool | No | False | If True, add a numeric prefix to each item (e.g., "1. Hello") |
-
 ```python
+from langchain_dev_utils.message_convert import format_sequence
 text = format_sequence([
     "str1",
     "str2",
@@ -178,19 +133,6 @@ Includes the following features:
 
 `has_tool_calling` and `parse_tool_calling` are used to check and parse tool calls.
 
-**`has_tool_calling` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `message` | AIMessage | Yes | - | AIMessage object |
-
-**`parse_tool_calling` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `message` | AIMessage | Yes | - | AIMessage object |
-| `first_tool_call_only` | bool | No | False | Whether to only check the first tool call |
-
 ```python
 import datetime
 from langchain_core.tools import tool
@@ -218,7 +160,7 @@ if has_tool_calling(response):
 Both can accept a `handler` parameter for custom breakpoint return and response handling logic.
 
 ```python
-from langchain_dev_utils import human_in_the_loop
+from langchain_dev_utils.tool_calling import human_in_the_loop
 from langchain_core.tools import tool
 import datetime
 
@@ -240,14 +182,7 @@ Includes the following features:
 
 #### 4.1 Agent Factory Functions
 
-In LangChain v1, the officially provided `create_agent` function can be used to create a single agent, where the model parameter supports passing a BaseChatModel instance or a specific string (when passing a string, it is limited to the models supported by `init_chat_model`). To extend the flexibility of specifying models via strings, this library provides a functionally identical `create_agent` function, allowing you to directly use models supported by `load_chat_model` (requires prior registration).
-
-**`create_agent` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `model` | str \| BaseChatModel | Yes | - | Model name or model instance. Can be a string identifier for a model registered with `register_model_provider` or a BaseChatModel instance. |
-| Other parameters | Various | No | - | All other parameters are the same as in `langchain.agents.create_agent` |
+In LangChain v1, the official `create_agent` function can be used to create a single agent; its `model` parameter accepts either a BaseChatModel instance or a specific string (when a string is provided, only models supported by `init_chat_model` are allowed). To extend the flexibility of specifying models via string, this library provides an equivalent `create_agent` function that lets you designate any model supported by `load_chat_model` (registration required beforehand).
 
 Usage example:
 
@@ -262,24 +197,26 @@ print(response)
 
 #### 4.2 Middleware
 
-Provides some commonly used middleware components. Below are examples of `SummarizationMiddleware` and `PlanMiddleware`.
+Provides some commonly used middleware components. Below, we illustrate with `ToolCallRepairMiddleware` and `PlanMiddleware`.
 
-`SummarizationMiddleware` is used for agent summarization.
+`ToolCallRepairMiddleware` is used to repair `invalid_tool_calls` generated by large language models.
 
 `PlanMiddleware` is used for agent planning.
 
 ```python
 from langchain_dev_utils.agents.middleware import (
-    SummarizationMiddleware,
+    ToolcallRepairMiddleware,
     PlanMiddleware,
 )
 
-agent=create_agent(
+agent = create_agent(
     "vllm:qwen3-4b",
     name="plan-agent",
-    middleware=[PlanMiddleware(), SummarizationMiddleware(model="vllm:qwen3-4b")]
+    middleware=[ToolCallRepairMiddleware(), PlanMiddleware(
+        use_read_plan_tool=False
+    )]
 )
-response = agent.invoke({"messages": [{"role": "user", "content": "Give me a travel plan to New York"}]}))
+response = agent.invoke({"messages": [{"role": "user", "content": "Give me a travel plan to New York"}]})
 print(response)
 ```
 
@@ -287,29 +224,14 @@ print(response)
 
 ### 5. **State Graph Orchestration**
 
-Includes the following features:
+Includes the following capabilities:
 
 - Sequential graph orchestration
 - Parallel graph orchestration
 
 #### 5.1 Sequential Graph Orchestration
 
-Sequential graph orchestration:
-Uses `create_sequential_pipeline`, supported parameters:
-
-**`create_sequential_pipeline` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `sub_graphs` | List[StateGraph] | Yes | - | List of state graphs to combine (must be StateGraph instances) |
-| `state_schema` | type | Yes | - | State Schema for the final generated graph |
-| `graph_name` | str | No | - | Name of the final generated graph |
-| `context_schema` | type | No | - | Context Schema for the final generated graph |
-| `input_schema` | type | No | - | Input Schema for the final generated graph |
-| `output_schema` | type | No | - | Output Schema for the final generated graph |
-| `checkpoint` | BaseCheckpointSaver | No | - | LangGraph persistence Checkpoint |
-| `store` | BaseStore | No | - | LangGraph persistence Store |
-| `cache` | BaseCache | No | - | LangGraph Cache |
+Use `create_sequential_pipeline` to orchestrate multiple subgraphs in sequential order:
 
 ```python
 from langchain.agents import AgentState
@@ -324,25 +246,25 @@ register_model_provider(
     base_url="http://localhost:8000/v1",
 )
 
-# Build sequential pipeline (all sub-graphs execute sequentially)
+# Build a sequential pipeline (all subgraphs executed in order)
 graph = create_sequential_pipeline(
     sub_graphs=[
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_time],
-            system_prompt="You are a time query assistant, can only answer the current time. If the question is unrelated to time, please directly answer that you cannot answer.",
+            system_prompt="You are a time-query assistant. You can only answer questions about the current time. If the question is unrelated to time, respond with 'I cannot answer that.'",
             name="time_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_weather],
-            system_prompt="You are a weather query assistant, can only answer the current weather. If the question is unrelated to weather, please directly answer that you cannot answer.",
+            system_prompt="You are a weather-query assistant. You can only answer questions about the current weather. If the question is unrelated to weather, respond with 'I cannot answer that.'",
             name="weather_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_user],
-            system_prompt="You are a user query assistant, can only answer the current user. If the question is unrelated to user, please directly answer that you cannot answer.",
+            system_prompt="You are a user-query assistant. You can only answer questions about the current user. If the question is unrelated to the user, respond with 'I cannot answer that.'",
             name="user_agent",
         ),
     ],
@@ -355,51 +277,36 @@ print(response)
 
 #### 5.2 Parallel Graph Orchestration
 
-Parallel graph orchestration:
-Uses `create_parallel_pipeline`, supported parameters:
-
-**`create_parallel_pipeline` Parameters:**
-
-| Parameter | Type | Required | Default | Description |
-|-----------|------|----------|---------|-------------|
-| `sub_graphs` | List[StateGraph] | Yes | - | List of state graphs to combine |
-| `state_schema` | type | Yes | - | State Schema for the final generated graph |
-| `branches_fn` | Callable | Yes | - | Parallel branch function, returns a list of Send objects to control parallel execution |
-| `graph_name` | str | No | - | Name of the final generated graph |
-| `context_schema` | type | No | - | Context Schema for the final generated graph |
-| `input_schema` | type | No | - | Input Schema for the final generated graph |
-| `output_schema` | type | No | - | Output Schema for the final generated graph |
-| `checkpoint` | BaseCheckpointSaver | No | - | LangGraph persistence Checkpoint |
-| `store` | BaseStore | No | - | LangGraph persistence Store |
-| `cache` | BaseCache | No | - | LangGraph Cache |
+Use `create_parallel_pipeline` to orchestrate multiple subgraphs in parallel:
 
 ```python
 from langchain_dev_utils.pipeline import create_parallel_pipeline
 
-# Build parallel pipeline (all sub-graphs execute in parallel)
+# Build a parallel pipeline (all subgraphs executed concurrently)
 graph = create_parallel_pipeline(
     sub_graphs=[
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_time],
-            system_prompt="You are a time query assistant, can only answer the current time. If the question is unrelated to time, please directly answer that you cannot answer.",
+            system_prompt="You are a time-query assistant. You can only answer questions about the current time. If the question is unrelated to time, respond with 'I cannot answer that.'",
             name="time_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_weather],
-            system_prompt="You are a weather query assistant, can only answer the current weather. If the question is unrelated to weather, please directly answer that you cannot answer.",
+            system_prompt="You are a weather-query assistant. You can only answer questions about the current weather. If the question is unrelated to weather, respond with 'I cannot answer that.'",
             name="weather_agent",
         ),
         create_agent(
             model="vllm:qwen3-4b",
             tools=[get_current_user],
-            system_prompt="You are a user query assistant, can only answer the current user. If the question is unrelated to user, please directly answer that you cannot answer.",
+            system_prompt="You are a user-query assistant. You can only answer questions about the current user. If the question is unrelated to the user, respond with 'I cannot answer that.'",
             name="user_agent",
         ),
     ],
     state_schema=AgentState,
 )
+
 response = graph.invoke({"messages": [HumanMessage("Hello")]})
 print(response)
 ```
@@ -410,4 +317,4 @@ print(response)
 
 - [GitHub Repository](https://github.com/TBice123123/langchain-dev-utils) — Browse source code, submit Pull Requests
 - [Issue Tracker](https://github.com/TBice123123/langchain-dev-utils/issues) — Report bugs or suggest improvements
-- We welcome contributions in all forms — whether code, documentation, or usage examples. Let's build a more powerful and practical LangChain development ecosystem together!
+- We welcome contributions in all forms — whether code, documentation, or usage examples. Let's build a more powerful and practical LangChain development ecosystem together!