This project is a Python application that uses the FastMCP library to create an MCP server. It also uses the openai library to interact with the OpenAI API.

• Python 3.6 or higher
• An OpenAI API key

1. Clone the repository:

   git clone <repository_url>
   cd try_fastmcp

2. Create a virtual environment:

   python -m venv env

3. Activate the virtual environment:

   env\Scripts\activate

4. Install the dependencies:

   pip install -r requirements.txt

5. Set the OpenAI API key:

   rename .env.example file to .env and fill your OPENAI_API_KEY

To run the MCP server, execute the following command:

python server.py

To run the MCP client, execute the following command:

streamlit run client.py

This MCP server provides the following tools:

• get_all_datasets: Retrieves all public datasets from the data.gouv.ci catalog.
• websearch_newssearch: Fetches news articles based on a user's query.
• add: Adds two numbers.
• subtract: Subtracts two numbers.
• multiply: Multiplies two numbers.
• divide: Divides two numbers.
• power: Calculates the power of two numbers.
• sqrt: Calculates the square root of a number.
• cbrt: Calculates the cube root of a number.
• factorial: Calculates the factorial of a number.
• log: Calculates the logarithm of a number.
• remainder: Calculates the remainder of two numbers division.
• sin: Calculates the sine of a number.
• cos: Calculates the cosine of a number.
• tan: Calculates the tangent of a number.

It also provides the following resource:

• greeting://{name}: Get a personalized greeting.

• Deepseek

• Add a "memory trimming" mechanism to limit the history length
• Implement conversation summarization for long exchanges
• Add an option to reset the conversation
• Save history in persistent storage (database, file)
• Format responses in Markdown