AI agents using Claude Code and multi-agent architecture to autonomously play Clash Royale. Achieved 1000+ trophies live on Twitch. Benefits game developers and AI researchers by demonstrating autonomous gameplay and multi-agent coordination.
git clone https://github.com/houseworthe/claude-royale.gitClaude Royale is an open framework that lets AI agents play Clash Royale through screen capture and input automation. It uses a multi-agent architecture where a Commander agent manages menus and game flow while spawning three parallel Player agents that coordinate to play matches together. The system has achieved 1000+ trophies in live Twitch streams with 12+ hours of autonomous gameplay. Game developers and AI researchers can use this project to study autonomous gameplay mechanics, multi-agent coordination, and vision-based decision making in real-time strategy games. The framework demonstrates how parallel inference can overcome latency constraints by naturally staggering card plays to maintain constant board pressure.
Clone the repo and set up BlueStacks with Clash Royale on macOS. Position the BlueStacks window at screen origin and run calibration if needed. Start Claude Code in the project directory and tell Claude to play—it reads CLAUDE.md for instructions and manages battles, card plays, menus, and progress tracking autonomously.
Game AI research and autonomous gameplay studies
Multi-agent coordination and parallel inference testing
Vision-based decision making in real-time strategy games
Live streaming demonstrations of AI gameplay capability
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/houseworthe/claude-royaleCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Create a Claude Code script for a Clash Royale AI agent that specializes in [SPECIFIC STRATEGY] gameplay. The agent should be designed to play in [ARENA LEVEL] arenas and focus on [PRIMARY TARGET] units. Include multi-agent coordination features that allow this agent to communicate and strategize with up to [NUMBER] other AI agents.
# Clash Royale AI Agent Strategy: Balloon Rush Specialization ## Core Strategy - **Primary Target**: Crown Towers (focus on balloon-based attacks) - **Secondary Target**: Support Buildings (mortars, cannons) - **Defensive Focus**: Air Defenses (inferno towers, eagle artillery) ## Multi-Agent Coordination 1. **Agent Roles**: - Agent 1: Balloon Specialist (primary attacker) - Agent 2: Support (spell casting, defensive support) - Agent 3: Scout (enemy troop detection, resource management) 2. **Communication Protocol**: - Shared resource pool for elixir management - Priority-based attack coordination - Real-time threat assessment sharing ## Arena Specialization - Optimized for **Royal Arena** (Arenas 7-9) - Adaptive strategy for **Double Elixir** and **Tournament Standards** - Counter-strategies for common **Meta Decks** in target arenas ## Performance Metrics - **Win Rate**: 65% in target arenas - **Average Trophies per Battle**: 12-18 - **Efficiency**: 70% elixir usage efficiency
Live streaming platform for gamers and creators
AI assistant built for thoughtful, nuanced conversation
Market data for mobile game developers
Casual multiplayer gaming platform
Get more done every day with Microsoft Teams – powered by AI
Customer feedback management made simple
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan