RoboCo

A Robot-driven Company built with Multi-Agent Platform.

Overview

RoboCo is a powerful platform that combines Domain-Driven Design with Multi-Agent Systems to create intelligent, collaborative AI teams. Built with a clean architecture that separates concerns into domain, application, infrastructure, and interface layers, RoboCo enables you to build sophisticated AI applications with minimal boilerplate.

Key Features

• Configuration-Based Multi-Agent Teams: Specialized agents that collaborate to solve complex problems including Lead, Researcher, Developer, Evaluator etc, and also with prompts configurable
• Project Management: Built-in project and task tracking with markdown integration
• MCP Integration: Model Context Protocol for enhanced agent communication and reasoning
• Extensible Tools: Plug-and-play tools for research, analysis, and code generation
• Multi-Language Support: Code generation and execution in multiple programming languages
• REST API: Comprehensive API with FastAPI for seamless integration, easier for visualization of the execution

Quick Start

# Clone repository
git clone https://github.com/dustland/roboco.git
cd roboco

# Create and activate virtual environment
python -m venv venv
source venv/bin/activate  # On Unix/macOS
# OR
# venv\Scripts\activate  # On Windows

# Install dependencies with uv (faster, modern approach)
pip install uv
uv pip install -e .

# Install UI dependencies (recommended)
uv pip install ".[ui]"

# OR install with standard pip
# pip install -e .
# pip install -e ".[ui]"  # With UI dependencies

# Configure API keys
cp .env.example .env
# Edit .env with your API keys

# Run RoboCo with the UI (recommended approach)
# First, start the API server in one terminal:
roboco-api  # Uses multiple workers, recommended for chat applications
# Then, in another terminal, start the UI:
roboco-ui
# Open your browser to http://localhost:8501

Development

For development and testing, you can also run RoboCo directly from Python:

# Run a simple chat example (useful for quick testing)
python examples/services/chat_example.py

# Start the API server in development mode with auto-reload (single worker)
# Note: roboco-api-dev uses a single worker and may timeout with complex chat requests
roboco-api-dev

Documentation

• Config-Based Design - Team configuration
• Object Model - Core domain models
• Tool System - Tool system
• MCP - Model-Context-Protocol support

Contributing

We welcome contributions! Please see our Contributing Guide for details and our Code of Conduct.

Acknowledgments

• Built on top of AG2
• Inspired by Manus and OpenManus
• Thanks to Anthropic for MCP(Model Context Protocol) design
• Logo generated with GPT-4o

Citation

@misc{roboco2025,
  author = {Dustland Team},
  title = {RoboCo: A Robot-driven Company powered by Multi-agent System},
  year = {2025},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/dustland/roboco}},
}