MCP Agent Tool Adapter

🚀 mcp-agent-tool-adapter

From protocol to intelligence: Powering agents with MCP.

🚀 Quick Start

☑️ Prerequisites

• Python 3.10+

⚙️ Installation

# Clone this repository
❯ git clone https://github.com/serkanyasr/mcp-agent-tool-adapter
❯ cd mcp-agent-tool-adapter

# Install dependencies
❯ pip install -r requirements.txt

✨ Features

• Modular tool invocation: Supports flexible invocation of multiple tools.
• Streaming interface: Provides efficient streaming data processing capabilities.
• Multi - language support: Adapts to different programming environments and frameworks.

📦 Installation

# Clone this repository
❯ git clone https://github.com/serkanyasr/mcp-agent-tool-adapter
❯ cd mcp-agent-tool-adapter

# Install dependencies
❯ pip install -r requirements.txt

💻 Usage Examples

Basic Usage

Run the Google ADK CLI agent:

❯ python app_client_adk.py

Run the LangGraph ReAct CLI agent:

❯ python app_client_langgraph.py

Ensure that your mcp_config.json defines the tools, for example:

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["@modular-protocol/filesystem"]
    }
  }
}

📚 Documentation

📍 Overview

The MCP Agent Tool Adapter enables modular tool invocation through the MCP protocol and provides agents based on Google ADK or LangGraph, allowing these agents to dynamically reason with tools.

This project transforms MCP tools into:

• 🤖 Google ADK agents with a streaming FastAPI or CLI interface
• 🧠 LangGraph agents using ReAct + streaming tool planning

🧱 Project Structure

mcp-agent-tool-adapter/
├── mcp_client/                # Core client implementation (modular)
│   ├── client.py              # MCPClient & MCPServer
│   ├── tool_loader.py         # High - level asynchronous loader
│   └── types.py               # Shared type definitions
├── app_client_adk.py         # Google ADK agent + CLI chat
├── app_client_langgraph.py   # LangGraph agent + ReAct CLI chat
├── mcp_config.json           # Example MCP tool configuration
├── requirements.txt          # Dependencies
└── README.md

🧠 Tool Architecture

MCP tools connect to the agent through MCPClient and handle:

• Tool server startup (via subprocess)
• Session management
• Tool adaptation (langgraph or Google ADK)
• Clean - up work

You can dynamically switch the agent type by changing the tool_type in the application to "google" or "langgraph".

🧪 Development and Testing

Project Process

1. Initialization: Configure the MCP server and load the required tools.
2. Agent startup: Run the Google ADK or LangGraph agent to receive requests.
3. Tool interaction: Dynamically invoke the corresponding tools for processing as needed.
4. Result feedback: Return the processed results to the client through the streaming interface.

Important Note

⚠️ Important Note

Ensure that all dependencies, including libraries such as fastapi and langchain, are installed. Correctly specify the tool paths and parameters in the configuration file to avoid runtime errors. When processing large - scale data, it is recommended to optimize network transmission and resource management to improve performance.

🔧 Technical Details

MCP tools connect to the agent through MCPClient. It takes charge of tool server startup (via subprocess), session management, tool adaptation (langgraph or Google ADK), and clean - up work. You can dynamically switch the agent type by changing the tool_type in the application to "google" or "langgraph".

📄 License

This project is licensed under the MIT License. For more information, please refer to the LICENSE file.