mcp-agent

Overview

The MCP Agent Orchestrator is a professional-grade Python implementation of the Model Context Protocol that provides a structured environment for LLMs to interact with external tools and knowledge bases. It features a decoupled client-server architecture using StdIO for communication, with specialized components including an intelligent client bridge, weather service server with real-time API integration, and a RAG knowledge server with vector database capabilities. The implementation strictly adheres to MCP specifications and supports multiple LLM providers including OpenAI and Qwen.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• get_weather

  Fetch real-time weather data from external REST APIs

  Try:What's the weather like in London?Try:Get current conditions for Tokyo
• query_knowledge_base

  Search ingested documents using vector embeddings

  Try:Find information about climate change in the documentsTry:What does the PDF say about renewable energy?
• ingest_document

  Add PDF or TXT documents to the knowledge base

  Try:Upload the quarterly earnings reportTry:Add the research paper to my knowledge base

Note: Inferred from system architecture and descriptions rather than explicit tool documentation

Comparable tools

Installation

{
  "mcpServers": {
    "mcp-agent": {
      "command": "python",
      "args": ["/path/to/mcp-agent/client.py"]
    }
  }
}

FAQ

What MCP servers does this project provide?

The project provides two main MCP servers: a weather service server for real-time weather data and a RAG knowledge server for document intelligence and knowledge retrieval.

Can this framework be used with other LLM providers besides OpenAI?

Yes, the framework supports OpenAI-compatible interfaces, including providers like Qwen and DashScope through configurable API endpoints and models.

Compare mcp-agent with