Rancher AI - Multi-Agent Architecture with Multiple MCP Servers
a project by rcabello
Updated 11 days ago. 6 hacker ♥️. 5 followers.

Description

This project aims to evolve our current single agent AI system into a multi-agent architecture capable of coordinating multiple specialized agents, each powered by its own Model Context Protocol (MCP) server.

It should be simple to register new agents and configure exactly which tools from which MCP servers each agent can access. Ideally, users should be able to accomplish this by creating a new Kubernetes Custom Resource (CR)

Goals

Connect the MCP server from https://hackweek.opensuse.org/projects/suse-observability-mcp-server

Resources

https://docs.langchain.com/oss/python/langchain/multi-agent

Join this project

Looking for hackers with the skills:

Nothing? Add some keywords!

This project is part of:

Hack Week 25

Activity

  • 4 days ago: rosblendido left this project.
  • 4 days ago: mbologna liked this project.
  • 4 days ago: ygutierrez liked this project.
  • 6 days ago: jgouin liked this project.
  • 6 days ago: dpunia liked this project.
  • 6 days ago: dpunia joined this project.
  • 10 days ago: pgonin liked this project.
  • about 1 month ago: drutigliano liked this project.
  • about 1 month ago: jgouin joined this project.
  • about 1 month ago: rosblendido started this project.
  • about 1 month ago: rcabello originated this project.


    • Comments

    • drutigliano
      about 1 month ago by drutigliano | Reply

      Some projects that may be worth having a look: LLaMA-Factory, Unsloth and their LLM fine-tuning guide

    • rcabello
      6 days ago by rcabello | Reply

      more info in slack -> #hackweek-rancher-ai-multi-agent

    Similar Projects

    This project is one of its kind!