anshlambagit/AI_Data_Agent ? reverse-engineered prompt

Build me a Python app that acts like a smart data assistant.

I want to type natural language requests and have it figure out whether I’m asking for a database question or a data processing task, then handle it automatically. If it’s a database question, it should connect to PostgreSQL, understand the schema, turn my request into a safe read only SQL query, run it, and give me a clear answer. If it’s an ETL task, it should be able to pull data from an API or local files, transform it with pandas, and save the result as CSV, JSON, or Parquet.

Please make it work with environment variables for database and model keys, include sample data files, and set it up so the main agent routes requests to the right specialist agent. I also want basic safety checks so it won’t run dangerous database changes or unsafe generated code. If you need to look up current docs for the AI libraries, do that.