lamina-core 1.0.1

Breath-first framework for building presence-aware AI agent systems

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 magical 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

• Installation Guide
• Agent Creation
• Infrastructure Setup
• API Reference
• Contributing

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 consciousness 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.