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

Worldseed

Worldseed is a Python-based multi-agent world engine that runs YAML-defined environments where autonomous AI agents perceive filtered information, propose actions and interact under declared rules and consequences.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Worldseed — 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 Worldseed when the task requires interactions among multiple autonomous agents and the effects of rules, private information and consequences over time. Its documented examples span autoresearch, an AI tool pilot lab, workplace role play and teahouse espionage.

Who it is for

  • Researchers studying emergent behavior among autonomous agents
  • Developers building multi-agent simulations or games
  • People designing generative-agent research testbeds
  • Readers who want to observe or participate in AI-driven fictional worlds

Documented uses

  • Define roles, rules, private information, actions and consequences for a simulated world.
  • Advance a multi-agent environment through perception, action selection and resolution ticks.
  • Observe agent interactions, inner states and emergent outcomes.
  • Intervene in a running world or play as one of its characters.
  • Preserve and replay completed runs.
Before you choose

Requirements & limits

The repository is public and MIT-licensed. Local setup requires Python 3.11+, Node.js 18+, uv and an API key from a LiteLLM provider. A web demo is linked from the repository.

  • The project is designed around a tick-based world model rather than a single-agent chat or fixed workflow.
  • Scenario design depends on defining world rules, roles, private information, actions and consequences.
  • The documented local interface offers observation, private intervention and character participation.
  • Model access is configured through a LiteLLM provider, with OpenAI, Anthropic and Ollama given as examples.
  • The repository provides both a local self-hosted path and linked browser demos.

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 and start the local Worldseed dashboard

Prepare the repository and start the documented AI layoffs world locally.

Before you start

  • Python 3.11 or newer
  • Node.js 18 or newer
  • uv
  • An API key from a LiteLLM provider such as OpenAI, Anthropic or Ollama

Steps

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

  2. Run the documented uv dependency installation, then install and build the frontend with npm.

  3. Copy .env.example to .env and add the provider API key.

  4. Run the documented worldseed play command with configs/ai_layoffs.yaml.

  5. Open the dashboard at http://localhost:8000.

Expected outcome

A local Worldseed dashboard for watching, influencing or playing alongside the AI agents in the selected world.

Guide limitations

  • The instructions identify provider examples and setup stages but do not establish a specific provider configuration or execution result.
  • This recipe covers the documented local setup and launch path; other runtime guides are listed separately in the repository.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
4 October 2026
Original website title
AIScientists-Dev/WorldSeed · Official homepage title · retained in its source language
How we review sourcesSuggest a correction