Agentic MCP Server

MCP All-in-One Server

This is a versatile MCP server that demonstrates multiple capabilities of the Model Context Protocol (MCP). It provides tools for calculation, external integrations (webhooks), resources for context, and prompt templates.

Capabilities

🧮 Tools

Functions that the AI can execute:

• Calculator: Basic arithmetic operations (add, subtract, multiply, divide).
• Webhook Integration: send_to_webhook - Sends a text prompt to a configured n8n workflow and returns the response.

📚 Resources

Contextual data the AI can read:

• support://playbook: A full Customer Support Playbook containing company overview, tone rules, product details, and standard operating procedures.

📝 Prompts

Templates to structure AI interactions:

• webinar_to_blog: A structured prompt template to convert webinar transcripts into engaging blog posts.

Prerequisites

• Python 3.10+
• uv: A fast Python package installer and resolver.
  • Windows: powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
  • macOS/Linux: curl -LsSf https://astral.sh/uv/install.sh | sh
• Node.js & npm (optional, for the MCP Inspector)

Setup

1. Open your terminal in this directory (calculator_server).
2. Initialize the project (if you haven't already):

   uv init
   

3. Add the required dependencies:

   uv add "mcp[cli]" httpx
   

Running the Server

You can run the server directly using uv:

uv run server.py

Note: This will use stdio for communication, so it will expect input from stdin and won't show much output in the console unless connected to an MCP client.

Testing with MCP Inspector

To test the server using the MCP Inspector (a web-based tool to interact with your server):

npx @modelcontextprotocol/inspector uv run server.py

This will launch the inspector in your browser. You can:

• Tools: Test the calculator and webhook tools.
• Resources: View the content of support://playbook.
• Prompts: Execute the webinar_to_blog template.

Integration with Claude Desktop

To use this server with Claude for Desktop:

1. Open your Claude Desktop configuration file:

   • Windows: %APPDATA%\Claude\claude_desktop_config.json
   • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

2. Add the server configuration. You can copy the content from claude_desktop_config.json in this directory:

   {
     "mcpServers": {
       "calculator": {
         "command": "uv",
         "args": [
           "--directory",
           "C:\\Users\\Your_Name\\Desktop\\mcps\\calculator_server",
           "run",
           "server.py"
         ]
       }
     }
   }
   

   Important: Make sure the path in --directory matches the actual absolute path to your calculator_server folder.

3. Restart Claude for Desktop.