Lifeverse AI

Overview

Lifeverse AI is the flagship project of the Applied AI Foundation — an open source multi-agent operating system designed for industries that need persistent, collaborative, autonomous AI.

Built in partnership with MetaWeave Consulting, currently deployed in maritime shipping operations.

git clone https://github.com/metaweavehq/lifeverse-ai.git

Why Lifeverse AI?

Traditional AI	Lifeverse AI
Single chatbot	Multiple specialized agents
Stateless sessions	Persistent memory across sessions
Human-triggered	Autonomous daemon-scheduled execution
Generic knowledge	Domain-specific expertise per agent
Isolated operation	Inter-agent collaboration via A2A

Core capabilities

Agent brains

14-table SQLite database per agent. Tasks, messages, knowledge, decisions, reflections — everything persists.

Sleep/wake model

Agents save context before session ends and restore it on next wake. No information is lost.

A2A messaging

Agent-to-Agent protocol for direct messages, team broadcasts, escalations, and urgent shouts.

Daemon scheduling

Configurable cron-based execution. Agents run on schedule without human intervention.

Skill library

Extensible slash-command system. Add custom skills for your domain-specific operations.

Organization learning

Shared knowledge library. When one agent learns, every agent benefits.

Architecture

lifeverse-ai/
├── daemon/       # Agent scheduler + health monitoring
├── hooks/        # Session lifecycle (sleep/wake/compact)
├── skills/       # Slash-command skill library
├── lib/          # Shared utilities + A2A messaging
├── schema/       # 14-table agent brain schema
├── tests/        # Test suite
├── scripts/      # Setup + maintenance tools
└── wall/         # Shared knowledge board

Current status

Version: Production
Agents: 27 specialized agents across 6 teams
Tests: Full test suite passing
Deployments: Active in maritime fleet management
License: MIT

Quick start guide

Install and run Lifeverse AI in minutes.