DREAMv1
Distributed Reasoning and Execution Architecture for Multi-agents
An open-source framework for building hierarchical multi-agent systems.
What is DREAMv1?
DREAMv1 is a framework for creating hierarchical multi-agent systems that can collaborate to solve complex problems. Built with a focus on practical applications and modularity, DREAMv1 provides the infrastructure needed to orchestrate multiple AI agents with different roles and capabilities.
The project aims to combine the best aspects of both symbolic and neural approaches to AI, creating systems that can reason, adapt, and execute complex tasks through coordinated agent interactions.
Key Features
- Hierarchical Architecture - Organize agents in functional hierarchies for efficient task delegation and management
- Flexible Communication - Built-in messaging protocols for structured agent interactions
- Memory Systems - Short and long-term memory mechanisms for knowledge persistence
- Monitoring & Reflection - Tools for system introspection and performance analysis
- Extensible Design - Easily integrate with external tools, APIs, and models
Architecture
Understand the system design, core components, and how they work together to form a cohesive multi-agent framework.
View Architecture →Quickstart
Get up and running with DREAMv1 quickly. Installation instructions and basic usage examples to help you get started.
Get Started →Examples
See DREAMv1 in action with practical examples ranging from simple agent communication to complex hierarchical systems.
View Examples →API Reference
Comprehensive documentation of all classes, methods, and parameters available in the DREAMv1 framework.
View API →System Overview
Master Control Program
The central orchestrator that manages agent lifecycle, communication, and system resources. Coordinates high-level activities and maintains system state.
Agent Hierarchy
A structured organization of agents with different roles and capabilities. Executive agents coordinate while specialist agents execute domain-specific tasks.
Communication Protocol
Standardized messaging system for agent interactions, supporting various message types, priorities, and delivery mechanisms.
Memory Systems
Working and long-term memory structures for storing and retrieving information across agent interactions and system sessions.
Knowledge Base
Centralized repository for storing structured knowledge, accessible to all agents for learning, reasoning, and decision-making.
Monitoring & Reflection
Tools for system self-assessment, performance tracking, and process improvement through continuous feedback loops.
Project Status
We are actively developing DREAMv1, continuously refining and enhancing its features. Your contributions are invaluable as we work towards achieving a stable and robust system. Check out our contribution guideline.
Get Involved
Hierarchical Multi-Agent Architecture
DREAMv1 implements a hierarchical structure where agents are organized in layers based on their roles and responsibilities. This design enables efficient task delegation, specialized expertise, and coordinated action.
Key Capabilities
- Hierarchical Coordination - Structured agent organization for efficient task handling
- Memory Systems - Knowledge persistence across agent interactions
- External Tool Integration - Seamless connection to APIs and services
- Flexible Communication - Structured message passing between agents
- Self-reflection - Performance monitoring and adaptation