M5_UGL_1_R - Plan execution - Extracted Answer shows: None

Course Q&A Agentic AI
,
1

Hello, the extracted answer output shows NONE. I have checked all all variables. The generated Python code creates the right answer. Does anyone have the same effect?

Thanks in advance!

Martin

2

Can you try printing the “result” instead of “result[“answer”]“ and check whether the agent is working correctly? if not could you share what the “result“ prints?

3

I had the same problem. The issue seems to come from the fact that the agent generates wrong python code:

cond = (
    (S.quantity_in_stock > 0) &
    (S.price < price_limit) &
    (S.description.test(lambda desc: shape_keyword in desc.lower(), desc))
)

This line is not correct:

(S.description.test(lambda desc: shape_keyword in desc.lower(), desc))

In order to fix it, I modified the agent prompt, and added an instruction:

Constraints:
- Use TinyDB Query for filt...
- Keep code clear an...

# THIS IS THE NEW INSTRUCTION
- Make sure the Query syntax is correct, especially when invoking description.test

With that, the code generated is different:

# Filter by style in description
if style_keyword:
    cond = Q.description.test(lambda desc, sk=style_keyword: sk in desc.lower())

And works as expected:

Plan Execution · Extracted Answer

Yes, we have our Classic round sunglasses in stock for $60.

Cheers

1 Like
4

@lesly.zerna can you verify this finding?

5

Thanks! I have rerun several times. Now it works.

Regards, Martin

2 Likes
6

Thanks for reporting and thanks @TMosh for the support!

We’ll also double check with the team!