Dev Assistant MCP

🧪

The Embedded Dev Assistant MCP server is currently in Beta. We're actively improving it based on developer feedback. If you have any feedback, please share it with your Technical Solutions representative or the product team.

The Embedded Dev Assistant Model Context Protocol (MCP) server enables AI-powered code editors like Cursor and Windsurf, as well as general-purpose tools like Claude Desktop, to interact directly with the Embedded Payroll API and relevant documentation.

The server can be used for AI-assisted development to ensure your environment has access to up to date documentation, and for prototyping features guided by available API functionality.

What is MCP?

Model Context Protocol (MCP) is an open standard that allows AI applications to securely access external data sources and tools. The Dev Assistant MCP server provides AI agents with:

• API reference access for searching and fetching endpoints, responses, etc.
• Documentation search capabilities to search and fetch guides
• Code and snippet generation capabilities

Setup

Gusto Embedded hosts a remote MCP server at https://embedded-payroll.readme.io/mcp. Configure your AI development tools to connect to this server.

{
  "mcpServers": {
    "embedded-payroll": {
      "url": "https://embedded-payroll.readme.io/mcp"
    }
  }
}

Testing your MCP setup

Once configured, restart your editor and test the connection:

1. Open your AI editor and start a new chat
2. Verify the connection by asking: "What MCP tools do you have access to?"
   • You should see the Embedded Payroll MCP server tools listed
3. Try these example prompts:
   • "How do I authenticate with the Gusto Embedded API? What are the best practices?"
   • "Show me the request body schema for creating a new employee"
   • "Generate a Python code snippet for running payroll"
   • "What webhooks are available for payroll events?"
   • "Show me an example of payroll reversals and how they work"
   • "Create a mock terminations UI using the Gusto Embedded API"

Expected behavior: Your AI assistant should reference Gusto's documentation directly and provide accurate, current information about the API.

Troubleshooting

Connection issues:

• Verify the MCP server URL is correct: https://embedded-payroll.readme.io/mcp
• Restart your IDE after updating the configuration file
• Check that your configuration file is valid JSON (use a JSON validator)

Tools not appearing:

• Ask your AI assistant "What tools do you have access to?" to confirm the MCP server loaded
• Check your IDE's MCP logs for connection errors
• Ensure your IDE supports SSE (Server-Sent Events) for MCP connections