Skip to main content

​
Overview

Use agent(name: Optional[str] = None, version: Optional[int] = None, method_name: Optional[str] = None) to mark an agent run. Tasks and tools executed inside the agent are captured beneath this span.

​
Function usage

from keywordsai_tracing.decorators import agent, task

@agent(name="support_agent")
def support_agent():
    @task(name="answer")
    def answer(question: str):
        return f"Answering: {question}"
    return answer("How to trace?")

print(support_agent())

​
Class usage

from keywordsai_tracing.decorators import agent, task

@agent(name="assistant", method_name="run")
class Assistant:
    @task(name="respond")
    def respond(self, input_text: str):
        return f"Echo: {input_text}"

    def run(self):
        return self.respond("Hello")

print(Assistant().run())

​
Parameters

  • name: agent display name; defaults to function/class name
  • version: optional integer version
  • method_name: when decorating a class, the method to wrap

​
Notes

  • Agent spans set the workflow name context for nested spans