MCP Employee API Server
A Model Context Protocol (MCP) server that provides tools for managing employee data through a REST API. This server exposes employee management operations as MCP tools that can be used by AI assistants and other MCP clients.
Features
- Employee Management: Full CRUD operations for employee data
- REST API Integration: Connects to a local employee API server
- MCP Protocol: Exposes functionality through the Model Context Protocol
- Async Operations: Built with async/await for optimal performance
- Error Handling: Robust error handling for API requests
Available Tools
The server provides the following MCP tools:
get_employees()- Retrieve all employeesget_employee(id)- Get a specific employee by IDadd_employee(name, age)- Create a new employeeupdate_employee(id, name, age)- Update an existing employeedelete_employee(id)- Delete an employee by ID
Prerequisites
- Python 3.13 or higher
- A running employee API server at
http://localhost:8000
Installation
-
Clone the repository:
git clone https://github.com/JoseGarayar/mcp_test.git cd mcp_test -
Clone the api employee repository:
git clone https://github.com/JoseGarayar/api_employees.git -
Install dependencies using uv:
uv sync
Usage
Running the MCP Server
Start the MCP server using stdio transport:
uv run python main.py
The server will run and listen for MCP protocol messages via stdin/stdout.
API Configuration
The server is configured to connect to a local API server at http://localhost:8000. You can modify the URL_BASE constant in main.py to point to a different API endpoint.
Example API Endpoints
The server expects the following API endpoints to be available:
GET /employees- List all employeesGET /employees/{id}- Get employee by IDPOST /employees- Create new employeePUT /employees/{id}- Update employeeDELETE /employees/{id}- Delete employee
Development
Project Structure
mcp_test/
├── main.py # Main MCP server implementation
├── pyproject.toml # Project configuration and dependencies
├── README.md # This file
└── uv.lock # Lock file for dependencies
Dependencies
httpx- Async HTTP client for API requestsmcp[cli]- Model Context Protocol implementation
Development Dependencies
ruff- Python linter and formatter
Error Handling
The server includes comprehensive error handling:
- Network timeouts (30 seconds)
- HTTP error status codes
- Invalid HTTP methods
- Connection failures
All errors are gracefully handled and return None for failed operations.
License
This project is part of a test implementation for MCP server development.