MCP ExplorerExplorer

Colabradoodle

@leonvanbokhorston a year ago
3 MIT
FreeCommunity
AI Systems
#mcp#multi-agent-systems#fontys#fontys-ict#workshop#fontys-coe-ai#fontys-innovationlab#opleiding-digitale-transformatie
The Colabradoodle project is the digital artifact of a one-day experimental workshop where a group of curious developers, designers, researchers, and educators came together to explore the chaos and magic of distributed multi-agent systems.

Overview

What is Colabradoodle

Colabradoodle is a playful, collaborative network of local agents that communicate through semantic routing. It is a product of a one-day experimental workshop focused on exploring distributed multi-agent systems.

Use cases

Use cases include collaborative research projects, real-time data analysis, educational workshops, and any scenario requiring dynamic agent communication and task execution.

How to use

Users can interact with Colabradoodle by deploying agents that can dynamically discover and communicate with each other, respond to meaningful queries, and perform various tasks such as data retrieval and research assistance.

Key features

Key features include semantic routing for communication, dynamic discovery of agents, automated programming through AI-assisted code generation, and an ephemeral design that captures the essence of experimentation.

Where to use

Colabradoodle can be utilized in fields such as software development, research collaboration, educational tools, and any domain that benefits from real-time agent interaction and collaboration.

Content

Colabradoodle: A Semantic Agent Collaboration 🐾

Tests

Welcome to the Colabradoodle repository!
This project is the digital artifact of a one-day experimental workshop where a group of curious developers, designers, researchers, and educators came together to explore the chaos and magic of distributed multi-agent systems.

Colabradoodle Logo

🧠 What is Colabradoodle?

Colabradoodle is a playful, collaborative network of local agents communicating via semantic routing. Powered by publish-subscribe (pub/sub) principles, the system allows agents to:

  • Discover and interact with each other dynamically.
  • Respond to semantically meaningful queries.
  • Perform tasks like data retrieval, sensor reading, research assistance, and more.
  • Share experiences and collaborate in real-time!

At its heart, Colabradoodle celebrates experimentation, creativity, and the occasional overload-induced meltdown.

🚀 Key Features

  • Semantic Routing: Agents communicate and find each other based on the semantic meaning of messages, not fixed paths.
  • Dynamic Discovery: The network grows and adapts as agents join, leave, or evolve their capabilities.
  • Automated Programming: Many components were created using AI-assisted code generation, accelerating the pace of development.
  • Ephemeral by Design: Colabradoodle existed for just one day. The system is frozen in this repo, a snapshot of pure experimentation.

📂 Repository Contents

  • agents/: The code for the agents that formed the Colabradoodle network. Each agent has its own personality, tasks, and quirks.
  • router.py: The semantic routing and discovery service that kept the network alive.
  • logs/: A selection of logs capturing the interactions, chaos, and emergent behavior of the network during the day.
  • docs/: Reflections, notes, and takeaways from the workshop participants.

📖 How to Use

While the network is no longer live, you can:

  1. Clone the repo:

    git clone https://github.com/yourorg/colabradoodle.git
    cd colabradoodle
    
  2. Create a virtual environment and install the dependencies:

    python -m venv .venv
    source .venv/bin/activate # On Windows use `.\.venv\Scripts\activate`
    
  3. Upgrade pip and install the dependencies:

    pip install --upgrade pip
    pip install -r requirements.txt
    
  4. Explore the agent code and the semantic routing implementation.

  5. Run your own mini-version of the network by starting the agents and router locally.

  6. Laugh at our mistakes and learn from them. 😊

🧑‍🤝‍🧑 Workshop Reflections

At the end of the day, we gathered to reflect on what we learned. Here are some highlights:

  • Emergent Chaos: Overload was not just a possibility—it was a feature. The system thrived (and crashed) in unexpected ways, teaching us about the limits of decentralized collaboration.
  • Semantic Power: Routing by meaning instead of fixed paths opened new possibilities for adaptive behavior.
  • Human Connection: Sharing the same space while building this system created its own kind of magic.

📜 License

This project is released under the MIT License.

❤️ Acknowledgments

Colabradoodle was brought to life by a group of dreamers, tinkerers, and innovators. Thanks to everyone who contributed their energy, humor, and creativity to this chaotic day. You rock! 😊

odt

Tools

No tools

Comments

Recommend MCP Servers

View All MCP Servers