general

AGI-Llama: AI-Powered Classic Sierra Game Engine

AGI-Llama is an AI-powered reimagining of Sierra's classic Adventure Game Interpreter engine that uses large language models to generate dynamic narratives and

Retro gaming enthusiasts modernize classic Sierra AGI games using natural language processing through AGI-Llama integration.

Installation Steps:

Clone the repository: git clone https://github.com/jalfonsosm/agi-llm

Configure local inference with llama.cpp or connect cloud APIs (OpenAI, Hugging Face, Groq)

Supported Backends:

  • llama.cpp: Runs Llama 3, Qwen, Gemma models locally
  • BitNet: Enables 1.58-bit quantized models
  • Cloud APIs: OpenAI, Hugging Face (hf.co), Groq endpoints

Gameplay Features:

Replace rigid “verb noun” commands with conversational input in any language

SDL3 GPU acceleration provides modern rendering for King’s Quest, Space Quest, Leisure Suit Larry

Integration:

Extract LLM logic library for ScummVM or similar interpreters

This approach eliminates parser frustration while preserving authentic AGI game mechanics, making 1980s adventures accessible to multilingual audiences.