Describe the bug
The git branch status chip added in #11938 renders noticeably larger than the other built-in chips in the terminal input area (working directory, git branch, git diff stats).
The chip builds its own container instead of reusing the shared UDI chip styling, which causes three inconsistencies:
- Doubled padding — 4px vertical / 8px horizontal, while standard chips use the shared
chip_container() helper with 2px / 4px (UDI_CHIP_VERTICAL_PADDING / UDI_CHIP_HORIZONTAL_PADDING).
- Oversized icon — the branch icon is hard-coded to 14px, while standard chips size their icon to the chip font size (~12px, see
render_udi_chip).
- Different font — it always uses the UI font family, while the neighboring diff-stats chip uses the monospace family outside of agent view.
To reproduce
- Open Warp in a git repository whose current branch tracks an upstream and is ahead of or behind it.
- Look at the context chips around the input box.
- The branch status chip (e.g.
main • ↓ 90) is visibly taller and more padded than the neighboring ± 0 diff-stats chip and the other chips.
Expected behavior
The branch status chip should have the same height, padding, border, and font as the other UDI chips, differing only in its content.
I contributed the original chip in #11938 and have a follow-up fix ready; opening this issue to link the PR against.
Describe the bug
The git branch status chip added in #11938 renders noticeably larger than the other built-in chips in the terminal input area (working directory, git branch, git diff stats).
The chip builds its own container instead of reusing the shared UDI chip styling, which causes three inconsistencies:
chip_container()helper with 2px / 4px (UDI_CHIP_VERTICAL_PADDING/UDI_CHIP_HORIZONTAL_PADDING).render_udi_chip).To reproduce
main • ↓ 90) is visibly taller and more padded than the neighboring± 0diff-stats chip and the other chips.Expected behavior
The branch status chip should have the same height, padding, border, and font as the other UDI chips, differing only in its content.
I contributed the original chip in #11938 and have a follow-up fix ready; opening this issue to link the PR against.