Feature request: configurable thinking-mode output (progress / tool calls)

Summary

Right now, the model’s reasoning output in PinchTab appears to be fixed to tool calls only.

I would like this to be configurable.

The reasoning stream itself contains both:

• the model’s current progress
• the concrete tool calls

But at the moment, PinchTab seems to expose only the tool-call part.

Requested capability

Please allow configuring which parts of the reasoning stream are exposed, for example:

• progress only
• tool calls only
• progress + tool calls

Why this matters

This would make agent execution easier to follow and debug.

Question

Would you consider supporting this?