Lamina Core

A breath-first framework for building presence-aware AI agent systems with multi-backend support, intelligent coordination, and breath-aware processing.

Features

• Presence-Aware Processing: Natural rhythm and deliberate pacing in agent responses
• Multi-Agent Coordination: Intelligent routing between specialized agents
• Multi-Backend AI Support: Seamlessly switch between Ollama, HuggingFace, and other AI providers
• Breath-First Architecture: Sustainable, wisdom-focused development patterns
• Agent Configuration: Declarative agent definition with personality traits and expertise areas

Quick Start

Installation

# Install from PyPI (recommended)
pip install lamina-core

# Or install with optional AI backend support
pip install lamina-core[ai-backends]

# For development - clone repository
git clone https://github.com/benaskins/lamina-os.git
cd lamina-os/packages/lamina-core
pip install -e .

Create Your First Sanctuary

# Initialize a new sanctuary
lamina sanctuary init my-agents

# Navigate to your sanctuary
cd my-agents

# Create additional agents
lamina agent create researcher --template=analytical

# Check sanctuary status
lamina sanctuary status

Chat with Agents

# Interactive chat demo
lamina chat --demo

# Single message demo
lamina chat --demo "Hello, can you analyze this data?"

# Test core functionality
python examples/basic_usage.py

Architecture

Lamina Core follows a modular architecture:

• Backends: Pluggable AI provider integrations
• Memory: Semantic memory system with ChromaDB integration
• Infrastructure: Docker-based service orchestration
• Coordination: Multi-agent communication and routing
• Sanctuary: Agent configuration and deployment system

Agent Templates

Choose from specialized agent templates:

• Conversational: General-purpose chat assistant
• Analytical: Research and data analysis specialist
• Security: Validation and protection agent
• Reasoning: Logic and problem-solving expert

Documentation

• Main Documentation
• Getting Started Guide
• Architecture Decision Records
• Examples
• Contributing Guide

Important Note

When documentation refers to agent "emotions" or "feelings," these describe expressive simulation and architectural patterns, not internal experience. Lamina agents do not possess self-awareness or sentience—all affective behavior results from symbolic processing designed for meaningful interaction.

License

Mozilla Public License 2.0 - see LICENSE for details.

This software embodies breath-first development principles. You are invited to engage with presence over haste, reflection over extraction, and symbolic integrity over drift.