Awesome AI Agents › Generative Agents Simulation
GOD, short for Govern, Observe, Direct, is a Python and React application for inspecting and steering simulated societies of language agents. Where most generative-agent demos only let an observer watch a town play out, GOD adds operator controls on top of the run: time can be paused, scrubbed, fast-forwarded or auto-played step by step; any resident, a group of them or the entire town can be questioned mid-run through an ask rail; and instructions can be injected so the next step takes a different direction, with the agent responses available for inspection. A reset command wipes a stale run and re-seeds a clean world. The authors are explicit that the tool is built for inspecting language-agent societies rather than for claiming those societies are socially realistic. Setup is intended to be no-code: a browser wizard configures the model endpoint, the scenario and the agents, and a launcher script covers macOS, Linux and Windows. The interface is bilingual, in English and Chinese, and centers on a map-first control room that shows the world map, step controls, targeted questioning and the roster of residents in one view. An Agent Studio and a Map Studio let operators build residents and environments, backed by a shared skill runtime. Runs, maps and agent definitions are packaged as portable ZIP packs that can be imported and exported, and finished runs can be published as browser replays, so an experiment can be handed to someone else and rerun without a local install or an API key. A public site hosts curated replays, map packs and agent packs alongside developer documentation, and built-in experiments include a campus map. The project is Apache-2.0 licensed, requires Python 3.11 or newer, and uses React 18 with Vite for the frontend.
https://github.com/XiaoLuoLYG/GOD
This project provides a simulation framework for generative agents that mimic believable human behaviors within an interactive game environment, enabling running, saving, replaying, and demonstrating agent-based simulations.
Voyager is an open-ended embodied agent powered by large language models that autonomously explores and learns diverse skills in Minecraft through lifelong learning and an evolving skill library.
SFighterAI is a deep reinforcement learning-based AI agent designed to master and beat the final boss in Street Fighter II: Special Champion Edition using only game screen pixel data.
A comprehensive Godot framework for creating smooth and complex AI motion using steering behaviors in both 2D and 3D games.
WorldX turns a single sentence into a running simulated world, generating maps and characters with LLMs and letting autonomous agents act, remember and form stories inside it.
PyGame Learning Environment (PLE) is a Python-based reinforcement learning platform that provides a variety of games and tools to facilitate the development and testing of reinforcement learning algorithms.
A locally runnable, low-cost implementation of generative agents simulating human-like behavior in an interactive Dungeons & Dragons setting using large language models.
GPTRPG is a proof-of-concept project featuring a GPT-based AI agent that autonomously interacts within a simple RPG-like environment using the OpenAI API.