Modelcontextprotocol Agent Example
What is Modelcontextprotocol Agent Example
ModelContextProtocol-Agent-Example is a demonstration application that utilizes the Model Context Protocol (MCP) to provide tools to a language model (LLM). It features a simple agent that operates in German, leveraging the MCP to enhance its capabilities.
Use cases
Use cases include creating a German-speaking virtual assistant, conducting automated research in German, and providing real-time information retrieval for German-speaking users.
How to use
To use the ModelContextProtocol-Agent-Example, first install the required dependencies using either ‘poetry install’ or ‘pip install -e .’. Start the MCP server with ‘python mcp_example/mcp_server.py’. Set the OpenAI API endpoint and key as environment variables, then launch the agent with ‘python -m mcp_example’.
Key features
Key features include a web search tool, a Wikipedia search tool, and a function to retrieve the current daytime. The agent communicates in German and utilizes OpenAI’s GPT-4o-mini for responses.
Where to use
This application can be used in fields such as natural language processing, AI-driven customer support, and educational tools where multilingual capabilities are beneficial.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Overview
What is Modelcontextprotocol Agent Example
ModelContextProtocol-Agent-Example is a demonstration application that utilizes the Model Context Protocol (MCP) to provide tools to a language model (LLM). It features a simple agent that operates in German, leveraging the MCP to enhance its capabilities.
Use cases
Use cases include creating a German-speaking virtual assistant, conducting automated research in German, and providing real-time information retrieval for German-speaking users.
How to use
To use the ModelContextProtocol-Agent-Example, first install the required dependencies using either ‘poetry install’ or ‘pip install -e .’. Start the MCP server with ‘python mcp_example/mcp_server.py’. Set the OpenAI API endpoint and key as environment variables, then launch the agent with ‘python -m mcp_example’.
Key features
Key features include a web search tool, a Wikipedia search tool, and a function to retrieve the current daytime. The agent communicates in German and utilizes OpenAI’s GPT-4o-mini for responses.
Where to use
This application can be used in fields such as natural language processing, AI-driven customer support, and educational tools where multilingual capabilities are beneficial.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Content
Model Context Protocol
Model Context Protocol (MCP) is a standardized way to provides tools to an LLM.
Application
This demo showcases a simple agent application (in german language) that uses the Model Context Protocol to provide tools to an LLM.
The tools are provided through a server which can be found at mcp_example/mcp_server.py.
These contain a web search, a wikipedia search and a function for the current daytime.
The MCP client is located at mcp_example/mcp_client.py and can connect to the server.
The agent, which uses the MCPClient to connect to the MCP server is located at mcp_example/mcp_agent.py.
It contains a german agent, which answers questions in german language using OpenAI GPT-4o-mini.
Requirements
Install the requirements:
#poetry
poetry install
#or with pip
pip install -e .
The MCP server needs to be started before the agent can be used. The server can be started with the following command:
python mcp_example/mcp_server.py
The agent uses the OpenAI API, for which the endpoint and api key need to be set as environment variables.
OPENAI_ENDPOINT=<your_openai_endpoint> OPENAI_API_KEY=<your_openai_api_key>
Afterwards, the agent can be started with the following command:
python -m mcp_example
Dev Tools Supporting MCP
The following are the main code editors that support the Model Context Protocol. Click the link to visit the official website for more information.
