LoL Roaster

English | 简体中文

A local proxy that exposes the League of Legends Live Client Data API to your LAN, featuring an AI Toxic Coach that roasts you in real-time when you die.

✨ Features

• 🔍 Auto Detection — Automatically detects the LoL game client
• 🔄 HTTP Proxy — Proxies all /liveclientdata/* endpoints to LAN
• 📡 WebSocket Push — Real-time game data via WebSocket
• 📊 Status Dashboard — Web UI for game status, players, and events
• 🤖 AI Toxic Coach — LLM-powered roasts with dynamic themes (powered by Google Gemini)
• 🗣️ Premium TTS — ElevenLabs V3 Dialogue API with 60+ emotion audio tags, or free Edge TTS
• 🎵 Web Audio Player — Browser-based player with progress bar, pause/resume, replay, and history
• 🌐 Multilingual — Chinese & English UI and roast generation

Quick Start

# Install dependencies
npm install

# Copy and configure environment
cp .env.example .env
# Edit .env with your API keys

# Start (development)
npm run dev

# Start (production)
npm start

Open http://localhost:8099 in your browser to see the dashboard.

Configuration

All settings are managed via .env:

# LLM (Google Gemini)
LLM_ENABLED=true
LLM_API_KEY=your_gemini_api_key
LLM_BASE_URL=https://generativelanguage.googleapis.com/v1beta/openai/
LLM_MODEL=gemini-2.0-flash

# TTS Provider: "edge" (free) or "elevenlabs" (paid, with emotion tags)
TTS_PROVIDER=elevenlabs
ELEVENLABS_API_KEY=your_elevenlabs_api_key
ELEVENLABS_VOICE_ID=5mZxJZhSmJTjL7GoYfYI  # Karo Yang (Chinese)

ElevenLabs Emotion Tags

When using ElevenLabs V3 Dialogue API, the AI coach can express emotions:

[laughs] [sighs] [sarcastic tone] [angry] [whispers] [shouts] ...

60+ tags are available including emotions, delivery styles, reactions, and pacing controls.

🤖 AI Toxic Coach

The coach triggers roasts based on in-game events:

Event	Trigger
💀 Death	Player dies
⚔️ Kill	Player gets a kill
💰 CS Gap	Poor CS performance
🐷 Teammate Death	Teammate dies
🐉 Objective	Dragon/Baron/Herald taken

Each roast uses a random theme (e.g., "disappointed parent", "suggest uninstall", "compare to minions") for variety.

API Endpoints

Endpoint	Description
GET /status	Service status
GET /liveclientdata/allgamedata	All game data
GET /liveclientdata/activeplayer	Active player
GET /liveclientdata/playerlist	Player list
GET /liveclientdata/eventdata	Game events
GET /liveclientdata/gamestats	Game statistics
GET /audio/:hash	Serve cached TTS audio
WS /ws	WebSocket real-time push

Package as CLI Executable

# Build for both Windows and Mac
npm run build

# Windows only
npm run build:win

# Mac only
npm run build:mac

Output in dist/ directory. Users can run the executable directly without Node.js.

�️ Desktop App (Electron)

Package as a desktop application with system tray and built-in browser window:

# Development
npm run electron:dev

# Build for Mac (.dmg)
npm run electron:build:mac

# Build for Windows (.exe installer)
npm run electron:build:win

# Build for both
npm run electron:build

Output in dist-electron/ directory.

�🚀 Automated Release (GitHub Actions)

git tag v0.2.0
git push origin v0.2.0

Pushing a tag triggers GitHub Actions to automatically build:

• CLI: lol-roaster-win.exe + lol-roaster-macos (via pkg)
• Desktop: .dmg (macOS) + .exe installer (Windows) (via electron-builder)

All artifacts are published on the GitHub Releases page.

How It Works

1. Polls https://127.0.0.1:2999/liveclientdata/allgamedata to detect game status
2. Proxies all API requests to the local LoL client
3. Pushes game data via WebSocket to connected browsers
4. AI Coach monitors events → generates roasts via LLM → converts to speech via TTS → streams audio to browser

License

MIT