[FEATURE] How to know which steps crew took to complete the goal given to crew

[zahidiqbalnbs]

Feature Area

Core functionality

Is your feature request related to a an existing bug? Please link it here.

Hi,

I’m using Crew.ai for an agent-based application, where I have multiple agents equipped with different tools. When I call crew.kickoff(), I’m currently unable to track or view the specific path that Crew follows to achieve the goal.

Thanks for your help!

Describe the solution you'd like

In LangGraph, we can access the entire state of the conversation in the end result object, including messages like HumanMessage, AIMessage, ToolCall, ToolMessage, and so on, providing a complete sequence of interactions and path it took to achieve the goal. I’d like to know how I can achieve something similar with Crew.ai, and access the sequence of actions along with the results given by crew.kickoff() function. I just get raw property containing only final response.

Describe alternatives you've considered

No response

Additional context

No response

Willingness to Contribute

Yes, I'd be happy to submit a pull request

[Vidit-Ostwal]

Set verbose = True, both at agents and crew level.

[zahidiqbalnbs]

@Vidit-Ostwal thank you for the quick response. The logs are currently only visible at the terminal level. What I need is to have them included alongside the final result so that I can use them for agent evaluation. How can I achieve that?

[Vidit-Ostwal]

Try using the event listeners. Thanks & Regards, *Vidit Ostwal,* ***@***.*** ***@***.***>*

…

On Sun, Aug 3, 2025, 10:41 PM Zahid Iqbal ***@***.***> wrote: *zahidiqbalnbs* left a comment (crewAIInc/crewAI#3268) <#3268 (comment)> @Vidit-Ostwal <https://github.com/Vidit-Ostwal> thank you for the quick response. The logs are currently only visible at the terminal level. What I need is to have them included alongside the final result so that I can use them for agent evaluation. How can I achieve that? — Reply to this email directly, view it on GitHub <#3268 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A2OQKVNTQDXQLAAKCKVL2AL3LY7DHAVCNFSM6AAAAACDAGCTA2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTCNBYGU3DONRSGI> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[mouramax]

@zahidiqbalnbs, I recommend adding a crucial tool for this debugging/optimization phase of your agentic system: a monitoring and observability platform. I suggest AgentOps or Phoenix, as they both offer smooth integration with CrewAI and a decent enough free tier.