EE309 Detective Game

An Android-based interactive detective game with LLM-powered dynamic content generation.

🎓 New to Kotlin/Android?

Start here: docs/GETTING_STARTED.md - Complete beginner-friendly guide with setup instructions!

Project Documentation

• docs/GETTING_STARTED.md: Complete beginner guide with setup instructions
• docs/PROJECT_BREAKDOWN.md: Detailed module breakdown with task descriptions, priorities, difficulties, and feasibility assessments
• docs/TASK_SUMMARY.md: Quick reference table with task dependencies and development phases
• docs/LLM_ARCHITECTURE.md: Detailed explanation of the Hybrid LLM Architecture approach
• docs/TECH_STACK.md: Tech stack documentation (matches actual build configuration)

Project Structure

The project is divided into 5 main modules:

1. Core Game Engine & State Management - Foundation for game state and flow
2. LLM Integration Layer (Hybrid Architecture) - Interface with LLM for content generation
   • LLM 1: Initial content generator (upfront generation)
   • LLM 2: Intro generator (introduction text)
   • LLM 3-4: Runtime generators (dialogue, descriptions)
   • Epilogue Generator: Game ending text
3. Game Content System - Management of characters, places, clues, and timeline
4. Android UI/UX Components - User interface and interaction
5. Game Logic & Validation System - Game rules and action handling

Tech Stack

Recommended Stack (see docs/TECH_STACK.md for details):

• Language: Kotlin
• UI: Jetpack Compose
• Architecture: MVVM + Hilt (DI)
• State Management: StateFlow
• Networking: Retrofit + OkHttp
• Database: Room Database
• LLM: Upstage API (Solar LLM) - Partnership support
• JSON: Gson (Retrofit) + kotlinx.serialization (Domain models)
• Async: Kotlin Coroutines