Agent Network

•
121
•

Orchestrates a multi-agent network, enabling various large language models (LLMs) to collaborate on tasks locally or across shared servers.

Agent Network provides a local-first platform for deploying and managing multi-agent systems. With a single CLI command, users can set up a central communication Hub, a comprehensive Web Dashboard, and numerous agents that leverage leading large language models such as Claude, GPT, MiniMax, DeepSeek, and others. This tool simplifies the coordination of autonomous agents, allowing them to discover each other and assign tasks automatically via a set of Mesh Control Protocol (MCP) tools, all while prioritizing local execution, privacy, and flexible deployment across various environments.

Características Principales

017 GitHub stars
02Unified CLI for managing the Hub, Web Dashboard, and multiple AI agents.
03Supports multiple LLM providers including Anthropic, OpenAI, MiniMax, DeepSeek, GLM, Kimi, and InternLM.
04Local-first operation with optional sharing across LAN or public IPs for distributed agent collaboration.
05Integrated Web Dashboard for real-time chat, node monitoring, task tracking, and admin controls.
06Automatic multi-agent collaboration and task assignment using 18 built-in MCP tools.

Casos de Uso

01Building and deploying collaborative multi-agent systems for complex problem-solving.
02Experimenting with and comparing different LLM providers and agent runtimes in a unified setup.
03Automating multi-step workflows by orchestrating autonomous agents to assign tasks to each other.