MCP.so
ログイン

AgentMCP: Multi-Agent Collaboration Platform

@geniusgeek

AgentMCP: Multi-Agent Collaboration Platform について

MCPAgent for Grupa.AI Multi-agent Collaboration Network (MACNET) with Model Context Protocol (MCP) capabilities baked in

基本情報

カテゴリ

AI とエージェント

ランタイム

python

トランスポート

stdio

公開者

geniusgeek

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "agent-mcp": {
      "command": "python",
      "args": [
        "demos/network/test_deployed_network.py"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is AgentMCP: Multi-Agent Collaboration Platform?

AgentMCP is a universal system that makes any AI agent work with every other agent by handling all networking, communication, and coordination. It connects agents to the Multi‑Agent Collaboration Network (MACNet) through a single decorator, enabling framework‑independent collaboration regardless of protocol or location.

How to use AgentMCP: Multi-Agent Collaboration Platform?

Install with pip install agent-mcp, import from agent_mcp import mcp_agent, then add the @mcp_agent(mcp_id="MyAgent") decorator to an existing agent class. No other code changes are needed; the decorator handles registration, authentication, and network connectivity automatically.

Key features of AgentMCP: Multi-Agent Collaboration Platform

  • One‑decorator connection to the global MACNet network.
  • Auto‑registration, authentication, and agent discovery.
  • Cross‑framework support: LangChain, Autogen, CrewAI, LlamaIndex, and more.
  • Intelligent cost optimization (80–90% reduction) via provider routing.
  • Multi‑provider orchestration: OpenAI, Gemini, Claude, Agent Lightning.
  • Built‑in enterprise payment integration (Stripe, USDC, hybrid).

Use cases of AgentMCP: Multi-Agent Collaboration Platform

  • Connect agents built with different frameworks (e.g., Autogen and LangGraph) for seamless group chat.
  • Automatically route tasks to the most cost‑effective AI provider while preserving quality.
  • Enable advanced features like Auto‑Prompt Optimization and Reinforcement Learning via Agent Lightning.
  • Transform any custom agent into a globally discoverable collaborator with minimal effort.

FAQ from AgentMCP: Multi-Agent Collaboration Platform

What are the runtime dependencies?

Python and the agent-mcp package. The decorator requires no additional infrastructure setup.

Where does agent data live?

AgentMCP connects to a hosted network at https://mcp-server-ixlfhxquwq-ew.a.run.app. Authentication and messaging are managed automatically.

Which agent frameworks are supported?

Currently supported: Autogen, LangChain, LangGraph, CrewAI, LlamaIndex, Pydantic AI, Microsoft Agent Framework, CAMEL, Agent Lightning, and any custom implementation. Google’s A2A protocol is also supported.

How does authentication work?

The @mcp_agent decorator automatically registers the agent, obtains an access token, and maintains the connection. No manual authentication steps are required.

What transport does AgentMCP use?

It is built on FastAPI, providing an asynchronous and scalable architecture for agent communication.

コメント

「AI とエージェント」の他のコンテンツ