While doing analysis of the console output from the crew, there is “Tool Input” and “Tool Output” shown but not seeing a way to define a path output for these to be captured and formatted. Is that available today in the 0.83.0 version?
What do you mean? What is your goal?
I’ve been using the output_file and output_log_file attributes for agents, tasks and crews. Issue is some of the more detailed context data shows in the console output but not in either of those files. Wanted to understand if there is a better way to control and get that detailed output to its own file. I would like to be able to save the content to then be ingested for different rendering and query options over time. Thanks.
Got it! You can force tool output as a result by setting the result_as_answer parameter to True. As stated in the docs:
This parameter ensures that the tool output is captured and returned as the task result, without any modifications by the agent.
Thanks! For this, I presume I do one per agent and update the task field to the name of the task the agent is executing?
In my case, these two examples for Agent and associated Task?
task_result = researcher.execute_task(research_task)
task_result = fact_checker.execute_task(fact_checker_task)
Figured it out. The key is that the task_result line has to go after all agents and tasks and before the crew is kicked off. I will do some testing to see if all the content is included in the output files and drop a note here. Are the output_file and output_log_file the only attributes needed to complete the print of this console output to a file?