Chain of thought modification

Hi guys, I have a question regarding the chain of thought process for tasks and agents. Every time I run my crew, I get this log:

lama3-vision | 
time=2025-02-12T16:20:20.381Z 
level=DEBUG 
source=routes.go:289 
msg="generate request" 
images=0 
prompt="<|start_header_id|>user<|end_header_id|>\n\n
### System:
\nYou are WhatsApp Data Manager. Experienced in extracting, processing, and storing data from various sources, with a focus on optimizing for retrieval and analysis.\nYour personal goal is: Efficiently manage and store WhatsApp chat data.\nYou ONLY have access to the following tools, and should NEVER make up tools that are not listed here:\n\n
Tool Name: 
call_whatsapp_endpoint\n
Tool Arguments: {'endpoint': {'description': 'WhatsApp API endpoint to call', 'type': 'str'}}\n
Tool Description: Tool to call WhatsApp API endpoints\n\nIMPORTANT: Use the following format in your response:\n\n\n
Thought: you should always think about what to do\n
Action: the action to take, only one name of [call_whatsapp_endpoint], just the name, exactly as it's written.\n
Action Input: the input to the action, just a simple JSON object, enclosed in curly braces, using " to wrap keys and values.\n
Observation: the result of the action\n\n\n
Once all necessary information is gathered, return the following format:\n\n\n
Thought: I now know the final answer\n
Final Answer: the final answer to the original input question\n\n\n
### User:\n\n
Current Task: Retrieve the list of WhatsApp chats with metadata (id, name, unreadCount) by calling the GET /chats endpoint.\n\nThis is the expect criteria for your final answer: JSON containing chat metadata (id, name, unreadCount).\nyou MUST return the actual complete content as the final answer, not a summary.\n\n
Begin! This is VERY important to you, use the tools available and give your best Final Answer, your job depends on it!\n\n
Thought:\n\n<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"

I want to modify the instructions to give the llm custom set of instructions on how to follow up on the different crews

Or is it best to finetune the model to change the prompt template?

You can change the agents prompts through prompt_template and system_template. Play with these to see what works for you Agents - CrewAI