Paychex MCP Server

This project is a demonstration MCP (Machine Callable Program) server that interacts with the Paychex Developer API documentation. It allows integration with various LLMs (Language Learning Models) to process queries against Paychex documentation.

Features

• Connect to Paychex Developer API documentation
• Support for multiple LLM providers (OpenAI, Anthropic, etc.)
• Environment-based configuration for secure key management
• RESTful API endpoints for querying Paychex data with LLMs

Setup

1. Clone this repository
2. Install dependencies

   npm install
   

3. Create a .env file based on .env.example

   cp .env.example .env
   

4. Edit the .env file with your API keys

Environment Variables

This project uses environment variables for configuration. Create a .env file in the root directory with the following variables:

# Server configuration
PORT=3000

# Paychex API keys
PAYCHEX_API_KEY=your_api_key_here
PAYCHEX_CLIENT_ID=your_client_id_here
PAYCHEX_CLIENT_SECRET=your_client_secret_here

# LLM configuration
LLM_TYPE=openai  # Options: openai, anthropic, azure, etc.
LLM_API_KEY=your_llm_api_key_here
OPENAI_API_KEY=your_openai_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here

Usage

Development

npm run dev

Production

npm start

API Endpoints

Get Paychex Data

GET /api/paychex/data?endpoint=optional_endpoint_path

Returns raw data from the Paychex API documentation.

Query with LLM

POST /api/paychex/query

Body:

{
  "query": "What are the available Paychex APIs?",
  "endpoint": "optional_endpoint_path"
}

Processes the query against Paychex documentation using the configured LLM.

Extending

Adding New LLM Providers

To add support for a new LLM provider:

1. Add a new service class in services/llm.js
2. Update the getLLMService() function to handle the new provider
3. Add the necessary environment variables to .env.example

Disclaimer

This is a demonstration project only. It is not officially affiliated with or endorsed by Paychex.