# Web World Models

> Bridge deterministic web code and generative LLMs via hybrid world models. State and physics defined by TypeScript code, LLMs generate narrative/aesthetics on top. Uses typed interfaces, deterministic hashing, graceful degradation—enabling scalable interactive environments from travel atlases to fictional worlds without databases.

- Skill: `adu2021/web-world-models` (Agent Skill)
- Install (CLI): `npx skillmds@latest add adu2021/web-world-models`
- Raw SKILL.md: https://api.skillmd.com/api/skills/adu2021/web-world-models/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: adu2021 (https://skillmd.com/u/adu2021)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/adu2021/web-world-models

---


## Overview

Web World Models combine structured code for invariant state with LLM-driven narrative generation.

## Core Technique

**Two-Layer Architecture:**

```python
# Physics layer (deterministic code)
class Physics:
    inventories: dict  # Code enforces consistency
    coordinates: XYZ

# Imagination layer (LLM-driven)
class Narrative:
    descriptions = llm.generate(physics_state)
    dialogue = llm.generate(context)
```

## References

- Physics layer as TypeScript code
- Imagination layer for aesthetics
- Deterministic hashing for consistency