function call can not get call_id

[goooodle]

return FunctionTool(
name=tool_name,
description=tool_desc,
params_json_schema=params_schema,
on_invoke_tool=create_run_function(service_key, service_name),
strict_json_schema=is_service_tool,
)

def create_run_function(service_key: str, service_name: str):
async def run_function(context, params_str: str) -> str:

  #当执行多个相同方法名,但是参数不同时,这里无法定位到是哪次调用, 因为该方法无法获取call_id
  #When multiple methods with the same name but different parameters are executed, it is impossible to 
      determine which call it is here because the method cannot obtain the call_id

    params = json.loads(params_str)

    try:
        result = await call_service(service_key=service_key, params=params)
       # context.context[call_id]["invoke_status"] = "success"
        return json.dumps(result, ensure_ascii=False)
    except Exception as e:
       # context.context[call_id]["invoke_status"] = "error"
        return f"Error executing tool: {str(e)}"

return run_function

Question

当执行多个相同方法名,但是参数不同时,这里无法定位到是哪次调用, 因为该方法无法获取call_id
When multiple methods with the same name but different parameters are executed, it is impossible to determine which call it is here because the method cannot obtain the call_id

[rm-openai]

Yeah, from the point of view of this SDK, multiple functions with the same name is an error. LLMs will generally have trouble with functions named the same; what's the use case?

[goooodle]

async def mock_execute(service_key, params:str):
    return '400'

def create_run_function(service_key: str, service_name: str):
    async def run_function(context, params_str: str) -> str:
        try:
           result = await mock_execute(service_key=service_key, params=params_str)
           # context.context[call_id]["invoke_status"] = "success"   // log process status, but cant not get call_id
           return result
        except Exception as e:
           # context.context[call_id]["invoke_status"] = "error"  // log process status, but cant not get call_id
          return f"Error executing tool: {str(e)}"
    return run_function


class FunctionArgs(BaseModel):
    product_name: str

async def main():
    tools = []
    tools.append(FunctionTool(
        name='query_price',
        description='query price',
        params_json_schema=FunctionArgs.model_json_schema(),
        on_invoke_tool=create_run_function('query_price', 'query price'),
        strict_json_schema=True,
    ))

    agent = Agent(
        name="gpt-4.1",
        instructions="You are a helpful assistant.",
        tools=tools
    )

    agent_result = await Runner.run(
        agent,
        input='query the prices of  macbook pro m1 and  macbook pro m2 ',
        context={})
    result = agent_result.final_output

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

The function query_price was called twice, with the following parameters: macbook pro m1 , Another parameter is: macbook pro m2，
I want to get the execution status of this function, but I can't get it because I can't get call_id in 'run_function' function

[rm-openai]

Ah got it. You want the call ID for some logging. Unfortunately we don't expose this right now. We could add it as an optional param to the on_invoke_tool - PR welcome!

[github-actions]

This issue is stale because it has been open for 7 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 3 days since being marked as stale.