Contextual/memory based answers for pandas agent is not accurate

[dark-art]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

#Methods Tried
##Append historical dialogues 
##Augment Query based on history (not so accurate sometimes)

##Append summary

Description

I am using deepseek model to generate pandas code from queries (pandas agent). If i add history context in the prompt, the pandas code is more of a hallucinated one. what can be best approaches for this?

System Info

custom pandas agent in langchain