---
description: 
globs: 
alwaysApply: true
---
project workflow:
In this project we are developing the talk to your data feature(prompt to SQL). From the Gradio Chat window (app.py), the langchain_mcp_client is launched, the client responds to incoming requests using the langchain ai-agent and postgres_mcp_server tools. the return consists of 4 parts: 1-result/data, 2-detail explanation of the result, 3-sql query used, and 4-chart of this result (pie-chart, bar-chart)).  
all these processes are reflected in the gradio chat window in a markdown format for the user.