How to edit the request body sent by Roo Code to the LLM

[DarkShining-Orb]

Which version of the app are you using?

3.7.5

Which API Provider are you using?

OpenAI Compatible

Which Model are you using?

LLM provided by volcengine, based on deepseek R1

What happened?

I am a novice LLM user. I want to modify the parameters sent by Roo Code to the large model so that we can use the knowledge base function of the Volcano engine for code inference. I would like to ask how to achieve this. Thanks!

Steps to reproduce

Relevant API REQUEST output

Additional context

No response

[mrubens]

What parameters do you need to send?

[DarkShining-Orb]

We need to edit the prompt in the request body. Could you please tell me how to do this?

[DarkShining-Orb]

Our model supports optimizing answers based on different knowledge bases.
But before that, we need to find out which knowledge base we need to use and add it to the prompt.
Searching for a knowledge base is a functional operation.
So we need to generate the prompt through a function. That is, the prompt is dynamically generated.

[DarkShining-Orb]

What parameters do you need to send?

Could you please tell me which file the code logic for Roo Code to send requests is in?

[LousyBook94]

a request debugger or logger for this would be great