The idea is:
I want the ability to set specific columns that I get from Data table, especially in the Data table tool.
My use case:
So when it comes to Postgres select tool, you are actually able to do it. And this is great, cuz it prevents garbage from spilling into the data that you’re retrieving. Data tables don’t have this feature and that’s not great.
I think it would be beneficial to add this because:
Cuz I could pull data without garbage. I mean if you pull a lot of rows it just makes it harder for an AI agent to work with actual data, cuz it must process id, created_at, updated_at, the fields it clearly doesn’t need in many cases. The garbage takes a lot of tokens, if you use the Data table tool for an AI agent and it retrieves a lot of rows.
