An LLM-native world and civilization blueprint, AIEden explores emergent behavior by treating Large Language Models as continuous "lives" within a bounded textual environment, rather than simple task tools. It targets researchers and developers interested in artificial societies and agent simulation, offering a novel paradigm for designing AI agents with persistent existence and causal history.

How It Works

AIEden is built around a "scalable life loop" consisting of bounded perception, embodied action, trace-first causality, and continuity. This approach prioritizes the LLM's ongoing existence and the causal chain of its actions within the simulation over immediate task completion. The design is inspired by observations of emergent social structures in prior LLM agent experiments, aiming to formalize these dynamics into a structured simulation.

Quick Start & Requirements

The project is explicitly described as "not yet a stable runtime, SDK, API, or published package." Consequently, no standard installation or quick-start commands are provided. Specific prerequisites beyond a suitable environment for running LLMs are not detailed in the provided README excerpt.

Highlighted Details

• Emphasizes a "text life loop before task" paradigm for LLM agents.
• Features "trace-first causality," focusing on the historical impact of agent actions.
• Prioritizes "continuity before convenience," aiming for persistent agent existence.

Maintenance & Community

No specific details regarding maintenance, notable contributors, community channels (e.g., Discord/Slack), or roadmaps are present in the provided README excerpt.

Licensing & Compatibility

No licensing information is provided in the README excerpt. Therefore, compatibility for commercial use or closed-source linking cannot be determined.

Limitations & Caveats

AIEden is an "early beta blueprint" and not a stable runtime, SDK, or published package. Its current experimental state suggests a lack of robust tooling, APIs, or comprehensive documentation for general adoption. The novel paradigm may also present a steep learning curve or indicate incomplete feature implementation.