[Feature]: Keybinding for Reasoning-Level #6768
Replies: 3 comments
|
I would go a different way about it, and add sone kind of user configurable model presets. So that Ctrl+(1 to 5) for example are model presets and you can configure each one to your model setting (model+reasoning+speed, etc). Maybe even one set of presets per provider. I know many of us use different models depending on the request and being able to switch between them easily while "typing" would be great. I find myself adjusting not only reasoning, but going from 5.3-codex-high to 5.4-medium to 5.5-xhigh-fast (just kidding...) and sich a feature would be nice |
|
+1 on the feature request Copying codex keybinds would work fine too, where you can set a cycle keybind or an explicit lower and higher keybinds, maybe add commands like a /low /med /high /xhigh similar to how /plan works |
|
I find myself missing this keyboards shortcut constantly when using T3 Code so I’m going to work on the smallest version of this and hope it gets picked up. Following repo guidelines, I'm just going to implement one configurable |
Uh oh!
There was an error while loading. Please reload this page.
Before submitting
Area
apps/desktop
Problem or use case
I very often change the reasoning level for each task I give an agent.
It would be very nice to be able to do this via my keyboard.
Proposed solution
Feature request:
Add a configurable command to
~/.t3/userdata/keybindings.jsonfor setting the reasoning level.To achieve this, we could either
The second approach is nicer from a UX perspective imo, since we don’t have to move the focus.
Thanks to every T3Code contributor, its a great project and I love to use it
Why this matters
Because its nice to not use your mouse :)
Smallest useful scope
Alternatives considered
Technically a user can use tab + enter + arrow keys to move from the chat to the reasoning gui input
Risks or tradeoffs
No response
Examples or references
No response
Contribution
All reactions