Bug caused by create_react_agent method and runnable asteam_events

[deershark]

Checked other resources

• This is a bug, not a usage question. For questions, please use GitHub Discussions.
• I added a clear and detailed title that summarizes the issue.
• I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example).
• I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue.

Example Code

import asyncio

from langchain_core.tools import tool
from langchain_openai import ChatOpenAI
from langgraph.prebuilt import create_react_agent


# Create an intentionally error-prone tool
@tool
def error_prone_tool(input: str) -> str:
    """A tool that always raises an error for demonstration purposes."""
    return 1 / 0


# Set up the agent
model = ChatOpenAI(model="gpt-4o-mini")
tools = [error_prone_tool]
agent = create_react_agent(model, tools)


async def demonstrate_issue():
    # Stream events to show missing tool_error event
    async for event in agent.astream_events(
        {"messages": "Please demonstrate a tool error"},
        version="v2",
    ):
        print(f"Event type: {event['event']}, Event name: {event['name']} Event data: {event['data']}")
        # Notice there's no 'tool_error' event type in the output

        # This is the issue - when a tool fails, we should get a 'tool_error' event
        # but currently we don't, making error handling difficult


if __name__ == "__main__":
    asyncio.run(demonstrate_issue())

Error Message and Stack Trace (if applicable)

Description

I am using the create_react_agent method to get all output events through astream_events. I want to get the error event to process and display the error information to the user.
If an error occurs in tool, there will be no on_tool_end event !!!!!
And, I see that the documentation of this method does not mention the error type of event (such as on_tool_error).
Is there any way to do this? If you do not use a custom callback.

Here is the output of running：

......
Event type: on_chain_start, Event name: tools Event data: {'input': {'messages': [HumanMessage(content='Please demonstrate a tool error', additional_kwargs={}, response_metadata={}, id='51b97b0a-2f8a-45c2-b6a8-a07f3c892882'), AIMessage(content='', additional_kwargs={'tool_calls': [{'index': 0, 'id': 'call_U8rGfLpsdQAoXhGZ01R4DpWd', 'function': {'arguments': '{"input":"Demonstrate tool error"}', 'name': 'error_prone_tool'}, 'type': 'function'}]}, response_metadata={'finish_reason': 'tool_calls', 'model_name': 'gpt-4o-mini-2024-07-18', 'system_fingerprint': 'fp_b705f0c291'}, id='run-00652a81-d2b9-469f-b413-79919c7c7839', tool_calls=[{'name': 'error_prone_tool', 'args': {'input': 'Demonstrate tool error'}, 'id': 'call_U8rGfLpsdQAoXhGZ01R4DpWd', 'type': 'tool_call'}], usage_metadata={'input_tokens': 53, 'output_tokens': 20, 'total_tokens': 66, 'input_token_details': {'audio': 0, 'cache_read': 0}, 'output_token_details': {'audio': 0}})], 'is_last_step': False, 'remaining_steps': 23}}

Event type: on_tool_start, Event name: error_prone_tool Event data: {'input': {'input': 'Demonstrate tool error'}}

Event type: on_chain_start, Event name: _write Event data: {'input': {'messages': [ToolMessage(content="Error: ZeroDivisionError('division by zero')\n Please fix your mistakes.", name='error_prone_tool', tool_call_id='call_U8rGfLpsdQAoXhGZ01R4DpWd', status='error')]}}
Event type: on_chain_end, Event name: _write Event data: {'output': {'messages': [ToolMessage(content="Error: ZeroDivisionError('division by zero')\n Please fix your mistakes.", name='error_prone_tool', tool_call_id='call_U8rGfLpsdQAoXhGZ01R4DpWd', status='error')]}, 'input': {'messages': [ToolMessage(content="Error: ZeroDivisionError('division by zero')\n Please fix your mistakes.", name='error_prone_tool', tool_call_id='call_U8rGfLpsdQAoXhGZ01R4DpWd', status='error')]}}
Event type: on_chain_start, Event name: _write Event data: {'input': {'messages': [ToolMessage(content="Error: ZeroDivisionError('division by zero')\n Please fix your mistakes.", name='error_prone_tool', tool_call_id='call_U8rGfLpsdQAoXhGZ01R4DpWd', status='error')]}}
......

System Info

System Information

OS: Linux
OS Version: langchain-ai/langgraph#1 SMP PREEMPT_DYNAMIC Fri Mar 8 11:32:16 CST 2024
Python Version: 3.11.5 (main, Sep 11 2023, 13:54:46) [GCC 11.2.0]

Package Information

langchain_core: 0.3.46
langsmith: 0.3.18
langchain_mcp_adapters: 0.0.5
langgraph_test: Installed. No version info available.

Optional packages not installed

langserve

Other Dependencies

httpx: 0.28.1
jsonpatch<2.0,>=1.33: Installed. No version info available.
langchain-core<0.4,>=0.3.36: Installed. No version info available.
langsmith-pyo3: Installed. No version info available.
langsmith<0.4,>=0.1.125: Installed. No version info available.
mcp<1.5,>=1.4.1: Installed. No version info available.
openai-agents: Installed. No version info available.
opentelemetry-api: Installed. No version info available.
opentelemetry-exporter-otlp-proto-http: Installed. No version info available.
opentelemetry-sdk: Installed. No version info available.
orjson: 3.10.15
packaging: 24.2
packaging<25,>=23.2: Installed. No version info available.
pydantic: 2.10.6
pydantic<3.0.0,>=2.5.2;: Installed. No version info available.
pydantic<3.0.0,>=2.7.4;: Installed. No version info available.
pytest: 8.3.4
PyYAML>=5.3: Installed. No version info available.
requests: 2.31.0
requests-toolbelt: 1.0.0
rich: Installed. No version info available.
tenacity!=8.4.0,<10.0.0,>=8.1.0: Installed. No version info available.
typing-extensions>=4.7: Installed. No version info available.
zstandard: 0.23.0

[vbarda]

Thanks for reporting - this is an issue in langchain-core, fix incoming