🤖 Pokemon LLM Agent

Autonomous AI Agent that plays Pokemon Red/FireRed using LLMs.

🌟 Key Features

• Visual Gameplay: Uses Vision Models (GLM-4V, GPT-4o) to "see" the game screen.
• Smart Navigation: A* Pathfinding with caching and cross-map routing.
• Narrative Engine: Generates live commentary and TTS audio.
• Chronicle Ecosystem: Mints achievements as Zora Coins on Base L2.
• 24/7 Autonomy: Resilient loop with error recovery and state persistence.

📚 Documentation

The full technical documentation is available at docs.llmletsplay.com.

🚀 Getting Started

• Installation Guide: Setup Python, mGBA, and ROMs.
• Cloudflare Setup: Deploy the Chronicle backend.
• Socials Setup: Configure Twitter & Discord bots.

🧠 Architecture

• System Overview: High-level diagrams.
• Stream Cycle: The main event loop timing.
• Pathfinding: Deep dive into navigation logic.
• Chronicle System: Token-gated content architecture.

🛠️ Guides

• Testing Runbook: End-to-end testing procedures.
• Dev Markers: How to annotate maps for the AI.

---

⚡ Quick Run

python run.py --mode ZAI --auto

🛠️ Development

Building Documentation

To build and serve the documentation site locally:

# Install dependencies
pip install -r requirements.txt

# Build and serve
python -m mkdocs serve

The site will be available at http://127.0.0.1:8000.