Getting error while calling a tool with two arguments

[YogeshSaini85]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

@tool
def get_schema_tool(args: str):
"""Can be used to get connections and objects associated with the applications and input is a python list
of applications e.g. ["app1","app2"]"""
app_list = ast.literal_eval(args)
return app_list

agent = create_react_agent(llm, tools, prompt)

agent_executor = AgentExecutor.from_agent_and_tools(
agent=agent,
tools=tools,
verbose=True,
handle_parsing_errors=True,
max_iterations=4000,
max_execution_time=60000,
return_intermediate_steps=True
)
agent_executor.agent.stream_runnable = False

agent_runnable = RunnableWithMessageHistory(
agent_executor,
lambda session_id: memory,
input_messages_key="input",
history_messages_key="chat_history",
)

agentrunnable.invoke({"input": user_input})

Error Message and Stack Trace (if applicable)

File "C:\Users\ysaini\OneDrive - Informatica\Desktop\ApiGenie\ConversationMemory\cai_copilot.py", line 103, in
output = process_input(u_input, conv_id)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\OneDrive - Informatica\Desktop\ApiGenie\ConversationMemory\cai_copilot.py", line 87, in process_input
out = copilot_agent.invoke({"input": user_input},
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 5343, in invoke
return self.bound.invoke(
^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 5343, in invoke
return self.bound.invoke(
^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 3022, in invoke
input = context.run(step.invoke, input, config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 5343, in invoke
return self.bound.invoke(
^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 4706, in invoke
return self._call_with_config(
^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 1923, in _call_with_config
context.run(
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\config.py", line 396, in call_func_with_variable_args
return func(input, **kwargs) # type: ignore[call-arg]
^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 4572, in _invoke
output = output.invoke(
^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\runnables\base.py", line 5343, in invoke
return self.bound.invoke(
^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain\chains\base.py", line 170, in invoke
raise e
File "C:\Users\ysaini\venv\Lib\site-packages\langchain\chains\base.py", line 160, in invoke
self._call(inputs, run_manager=run_manager)
File "C:\Users\ysaini\venv\Lib\site-packages\langchain\agents\agent.py", line 1629, in _call
next_step_output = self._take_next_step(
^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain\agents\agent.py", line 1337, in _take_next_step
for a in self._iter_next_step(
^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain\agents\agent.py", line 1420, in _iter_next_step
yield self._perform_agent_action(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain\agents\agent.py", line 1442, in _perform_agent_action
observation = tool.run(
^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\tools\base.py", line 688, in run
raise error_to_raise
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\tools\base.py", line 651, in run
tool_args, tool_kwargs = self._to_args_and_kwargs(tool_input)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\tools\base.py", line 574, in _to_args_and_kwargs
tool_input = self._parse_input(tool_input)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\ysaini\venv\Lib\site-packages\langchain_core\tools\base.py", line 509, in parse_input
input_args.model_validate({key: tool_input})
File "C:\Users\ysaini\venv\Lib\site-packages\pydantic\main.py", line 596, in model_validate
return cls.pydantic_validator.validate_python(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
pydantic_core._pydantic_core.ValidationError: 1 validation error for get_schema_tool
v__args
Input should be a valid list [type=list_type, input_value='["Salesforce", "ServiceNow"]', input_type=str]
For further information visit https://errors.pydantic.dev/2.9/v/list_type

Description

I have a tool which expects a string. This string is a string representation of a python list of strings. In the tool I do literal_eval to convert it into a python list.

But I get this error:
Input should be a valid list [type=list_type, input_value='["Salesforce", "ServiceNow"]', input_type=str]

Why is it expecting list when I have clearly defined the args as string.

System Info

System Information

OS: Windows
OS Version: 10.0.19045
Python Version: 3.12.6 (tags/v3.12.6:a4a2d2b, Sep 6 2024, 20:11:23) [MSC v.1940 64 bit (AMD64)]

Package Information

langchain_core: 0.3.5
langchain: 0.3.0
langsmith: 0.1.127
langchain_text_splitters: 0.3.0

Optional packages not installed

langgraph
langserve

Other Dependencies

aiohttp: 3.10.6
async-timeout: Installed. No version info available.
httpx: 0.27.2
jsonpatch: 1.33
numpy: 1.26.4
orjson: 3.10.7
packaging: 24.1
pydantic: 2.9.2
PyYAML: 6.0.2
requests: 2.32.3
SQLAlchemy: 2.0.35
tenacity: 8.5.0
typing-extensions: 4.12.2

[enzo-inc]

Hello,

According to the langhchain documentation on tool use, you can use the args_schema argument to the @tool decorator. This expects a pydantic BaseModel, which can have your string as its field. For example

class PydanticArgs(BaseModel):
    args: str

@tool(args_schema=PydanticArgs)
def get_schema_tool(args: str):
    """Can be used to get connections and objects associated with the applications and input is a python list
    of applications e.g. ["app1","app2"]"""
    app_list = ast.literal_eval(args)
    return app_list
    
input = '["app1", "app2"]'

get_schema_tool(input)

[YogeshSaini85]

Thank you @enzo-inc , it worked.