Openrouter: Error inside tool result when using functions without arguments

[philippHorn]

What happened?

Describe the bug
Using most models besides openai on openrouter does not work with functions that have no arguments.

To Reproduce
This script reproduces the issue, you just need to change the openrouter API key

import asyncio
from pprint import pprint

from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import (
    OpenAIChatCompletionClient,
)

import settings

# Define a model client. You can use other model client that implements
# the `ChatCompletionClient` interface.
model_client = OpenAIChatCompletionClient(
    model="anthropic/claude-3.5-sonnet",
    api_key=settings.OPENROUTER_KEY,
    base_url="https://openrouter.ai/api/v1",
    model_info={
        "vision": True,
        "function_calling": True,
        "json_output": False,
        "family": "claude-3.5-sonnet",
    },
)

def get_weather() -> str:
    """Get the weather for a given city."""
    return f"The weather is 73 degrees and Sunny."


# Define an AssistantAgent with the model, tool, system message, and reflection enabled.
# The system message instructs the agent via natural language.
agent = AssistantAgent(
    name="weather_agent",
    model_client=model_client,
    tools=[get_weather],
    system_message="You are a helpful assistant.",
)


# Run the agent and stream the messages to the console.
def main() -> None:
    result = asyncio.run(agent.run(task="What is the weather in New York?"))
    pprint(result)


main()

Result:

TaskResult(messages=[TextMessage(source='user', models_usage=None, content='What is the weather in New York?', type='TextMessage'),
                     ToolCallRequestEvent(source='weather_agent', models_usage=RequestUsage(prompt_tokens=389, completion_tokens=48), content=[FunctionCall(id='tooluse_crDs0nSRTziC4TS0Wg-JGg', arguments='', name='get_weather')], type='ToolCallRequestEvent'),
                     ToolCallExecutionEvent(source='weather_agent', models_usage=None, content=[FunctionExecutionResult(content='Error: Expecting value: line 1 column 1 (char 0)', call_id='tooluse_crDs0nSRTziC4TS0Wg-JGg', is_error=True)], type='ToolCallExecutionEvent'),
                     ToolCallSummaryMessage(source='weather_agent', models_usage=None, content='Error: Expecting value: line 1 column 1 (char 0)', type='ToolCallSummaryMessage')],
           stop_reason=None)

Expected behavior
The result of the tool call should be added to the agent messages

Which packages was the bug in?

Python AgentChat (autogen-agentchat>=0.4.0)

AutoGen library version.

Python 0.4.7

Other library version.

No response

Model used

anthropic/claude-3.5-sonnet

Model provider

OpenRouter

Other model provider

No response

Python version

3.10

.NET version

None

Operating system

None