[Feature Request] Claude Code model selection and performance transparency

Bug Description

You = Anthropic

I don't know if I can not call this model Fable anymore.
I have no idea what you've done to it, though I have a strong suspicion that it was done to push us, max people, to move to API usage since we are not good business for you.

This is sad. Really sad. You just didn't have to include it in the max plan.
Now, I won't be moving anytime soon to the API usage because I'm afraid I will find the same watered down Fable you serve us.
And I think I must not be the only one to think that, I am no genious.

Anybody remembers how Fable used to be ? The guy that given a prompt, it draws the architecture, comes up with tech specs, codes, tests and deploys in one go ? One shot from end to end ?
Was I dreaming that ? Anybody remembers this ?

Since your last update, "Fable" (and mind the "") became insufferable. It does not do what it's asked.
It enjoys a new sport which is going down the rabbit hole on every task, over complicate and over engineer things, guzzle tokens for 0 value since it needs thirty rounds to come to a result that's almost acceptable.
It writes too many explanations, and drown the gist in prose.

It became the guy in the office nobody wants to work with and always refuse his meetings invitations by default.
Do you see the picture ?

I am just venting here, I know Anthropic won't even read this, let alone do something about it.
This is exactly what hunger for money does to a beautiful product.
And it used to be the answer to GPT, how sad to see what became of "Fable".

Environment Info

• Platform: darwin
• Terminal: webstorm
• Version: 2.1.265
• Feedback ID: b0a488e4-5ffd-4e3c-9f6e-ce92c5ab0617

Errors

[]