Why don't you try to dispove me. Yes, they are _more_ deterministic than tool calls and consume less tokens.
How dared I to imply that some LLM output is more deterministic than the other, your LLM majesty. Shame on me and my entire family! For generations to come!
So sorry I implied that the code that doesn't work and has to be fixed later is deterministic in its execution and can be reused later instead of being re-generated from scratch!
Will I ever wash it off my name, your grace?
Anyway you are wrong, the tool calls are deterministic, generating a one-shot script isn't.
Calling scripts wins long term. Maybe you haven't had a chance to work with agents for months on the same task and observing them getting looped. That is your ignorance talking.
Why am I even discussing it with people who don't understand basic costs of time and money, ffs.