Recently, I’ve been revisiting one of my Android starter repositories, Android Booster, and modernizing it with two major goals in mind:
- Adopting a modular project structure
- Adding Claude Skills to support AI-assisted development workflows
The result is a cleaner, more scalable project that can serve as a production-ready starting point for new Android apps while also being optimized for modern AI coding assistants.
Modularization Support
One of the biggest updates to the project is the addition of full modularization support.
Android Booster is now organized into a core module with multiple submodules, each responsible for a specific layer of the application.
The project structure now includes:
- core-architecture – Base classes for activities, fragments, and ViewModels, along with architecture patterns such as MVI, MVVM, MVP, and VIPER. Includes StateFlow-based MVI contracts, dependency injection modules (Hilt), and an RxBus event system.
- core-ui – Shared UI components and helpers including RecyclerView base activities/fragments, BindingUtils (Coil image loading), Snackbar utilities, dialogs, display utilities, and unit helpers.
- core-network – Retrofit, OkHttp, and Gson configuration with a shared Retrofit interface and configurable base URL.
- core-database – Reserved for persistence layers such as Room.
- core-audio – Integration with HX Audio Player for music and sound playback.
- core-location – Location services support including modules for LocationManager and LocationRequest.
- core-common – Shared constants, connectivity helpers, SharedPreferences utilities, and other global configuration.
This structure aligns with recommended Android architectural guidance for scalable multi-module applications.
The goal is to make Android Booster useful as a drop-in starting architecture for new Android projects. Instead of repeatedly setting up dependency injection, networking layers, or shared UI utilities, those pieces are already structured and ready to use.
For anyone interested in learning more about modular Android architectures, the official Android documentation on modularization is an excellent reference.
The Android Booster repository can be found here: https://github.com/huhx0015/AndroidBooster
Claude Skills for AI-Assisted Development
Another interesting addition is Claude Skills support.
Android Booster now includes a claude/skills/ directory that provides structured guidance for AI agents on how to work with the repository.
These skills provide contextual information about:
- Architecture patterns used in the project
- Audio playback components
- Network configuration
- UI utilities and base classes
- Shared utilities and configuration
When the project is loaded in Claude Code or compatible tools, these skills help AI agents better understand how the project is structured and how to correctly use its components.
HXAudioPlayer Claude Skill
As part of this work, I also added a dedicated Claude Skill for HXAudioPlayer, which is used by Android Booster for audio playback.
The skill provides guidance to AI agents on how to correctly use the HXAudioPlayer components, including:
- HXMusic for music playback
- HXSound for sound effects
- lifecycle-aware audio handling
- recommended usage patterns
This allows AI coding assistants to generate correct usage examples automatically, making it easier to integrate audio playback into new features.
You can view the HXAudioPlayer Claude Skill here: https://github.com/huhx0015/HXAudioPlayer/tree/master/claude/skills/hxaudio-player