The Quality of output varies when used within the enchanted app vs when used as a completion tool in another app.

[sourcedexter]

I have tested the completion functionality for multiple tasks and I have noticed that a prompt yields better result within the enchanted app vs when it's used through completion in another application. Here's a detailed example of this scenario.

Scenario: Running a code generation task using the following prompt "python script to accept a web URL as an input parameter and parses the contents of that website and store the content into a text file"

1. Output in Enchanted:

2. Output in PyCharm Python Editor with temperature at 0.0 :

Observation: The prompt is not followed correctly. the part asking it to store data to file is missed entirely and formatting is off as well

3. Output in Pycharm Python Editor with temperature at 0.8:

Observation: The formatting of the code is completely messed up.

The completion intruction was : "Write a python script using the description below"

Overall no matter how i changed the completion instruction, the answer was quite bad compared to what I am able to get within the enchanted app.

I wanted to know if there is any way i can rectify this?