OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Multi-Agent Systems

Brunnfeld

Brunnfeld is a TypeScript research simulation in which up to 1,000 LLM agents participate in a rule-based medieval village economy across multiple villages.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Brunnfeld — OpenViST project profile in Multi-Agent Systems.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Brunnfeld when you want to run or adapt a multi-agent economy simulation and examine how LLM choices play out under deterministic rules.

Who it is for

  • Researchers studying multi-agent behavior under economic constraints
  • Developers exploring LLM-driven simulations or analysis tools
  • People interested in how scarcity, supply chains and markets shape agent decisions

Documented uses

  • Observe how agents respond to resource scarcity and market pressure.
  • Compare village economic outcomes across agent population sizes.
  • Study how explicit rules and agent choices interact in production, trade and governance.
Before you choose

Requirements & limits

The documented setup uses Node.js 18 or later and requires either an authenticated Claude Code CLI installation or an OpenRouter API key. OpenRouter mode streams tokens live; CLI mode does not.

  • The simulation separates LLM action choices from deterministic resolution of world rules.
  • Its economy includes linked production chains, a live order book, price indexing, spoilage, seasonal yields and debt.
  • The project supports up to 1,000 agents across as many as five villages.
  • A separate viewer build is described in the README.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up Brunnfeld with an LLM backend

Install the project and configure either the Claude Code CLI or OpenRouter as its LLM backend.

Before you start

  • Node.js 18 or later
  • An authenticated Claude Code CLI installation, or an OpenRouter API key
  • A local shell and npm

Steps

  1. Clone the Brunnfeld repository and change into its directory.

  2. Run npm install.

  3. Copy .env.example to .env and configure the chosen backend; leave OPENROUTER_API_KEY unset for CLI mode, or set it for OpenRouter mode.

  4. Run npm start.

Expected outcome

The simulation starts at tick 1 and streams to http://localhost:3333. OpenRouter mode streams agent tokens live; CLI mode displays responses after a full turn.

Guide limitations

  • The documented setup does not establish that a web viewer has been installed; its separate build steps are described in the README.
  • The CLI option requires Claude Code to be installed and authenticated.
Documentation 1 ↗Documentation 2 ↗
Evidence and transparency

Sources & review

Official repository ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
marcopatzelt/brunnfeld-agentic-world · Official homepage title · retained in its source language
How we review sourcesSuggest a correction