GitHub

Command Palette

Search for a command to run...

Quick Start

Fleet UI provides 2 installation tracks (Track A / Track B). This page helps you quickly choose "the right track for your team."

For installation/setup, proceed step-by-step at Start Installing with Install.


Common Prerequisites (Both Tracks)

Fleet UI is react-native-unistyles based, so Unistyles setup is required. That's why you use @fleet-ui/core which includes Fleet UI's base theme/tokens/configuration together.

Track A also installs @fleet-ui/core locally at project root. So you can modify it directly within the project if needed.


Which Track Should You Choose?

One-Line Track Definitions

  • Track A (Local Install): Use CLI to bring code into your project, then modify/override it "like your own code." (like Shadcn UI)
  • Track B (Package Install): Install via package manager, update through version upgrades.

How to Choose

Track A is easier to manage in most cases. Answer these questions—usually you'll have a conclusion in 1-2 steps.

  1. Do you need to frequently modify component/token implementations? → Yes: Track A / No: Go to next question.

  2. Is installation/update/CI stability your top priority? → Yes: Track B / No: Go to next question.

  3. Do multiple apps/packages need to share the same UI via versioning? → Yes: Track B / No: Go to next question.

  4. Is keeping code in the project (ownership) and quickly experimenting/modifying more comfortable? → Yes: Track A / No: Choose Track B first.


Quick Selection Criteria

Recommend Track A (Local Install, Source Ownership)

  • Need to quickly customize components/tokens
  • Want to directly change file structure/styles/behavior to match project conventions
  • Prefer "bringing UI base code to our project to manage" rather than "adopting an SDK"

Recommend Track B (Package Install)

  • Token overriding alone (e.g., radius/shadow/colors/typography) can match most styles
  • Installation/update/CI stability is top priority
  • Prefer using via theme/props combinations over code copying
  • Large team where "standard install → version upgrade" model is operationally advantageous

Next Steps