🔍 Rust MCP Server + Inspector Example: SSE Transport with an Add Tool

What is 🔍 Rust MCP Server + Inspector Example: SSE Transport with an Add Tool?

This project demonstrates how to set up an MCP (Model Context Protocol) server and client using Server-Sent Events (SSE) for communication. It includes a simple tool that adds two numbers and integrates with the RIG agent for LLM prompting.

How to use 🔍 Rust MCP Server + Inspector Example: SSE Transport with an Add Tool?

Clone the repository and run cargo run. In a separate terminal, start the MCP Inspector with npx @modelcontextprotocol/inspector sse http://127.0.0.1:3001/sse. Then open http://localhost:5173 in a web browser to view the Inspector UI.

Key features of 🔍 Rust MCP Server + Inspector Example: SSE Transport with an Add Tool

• Sets up a custom MCP server using ServerSseTransport
• Connects an MCP client to the server via SSE
• Registers a custom Add tool that adds two numbers
• Lists registered tools through MCP protocol
• Integrates with the RIG agent and OpenAI LLM

Use cases of 🔍 Rust MCP Server + Inspector Example: SSE Transport with an Add Tool

• Learning how to implement an MCP server with SSE transport
• Prototyping and testing custom MCP tools in a sandboxed environment
• Integrating an LLM agent with external tools via the MCP protocol
• Debugging MCP servers using the inspector web interface

FAQ from 🔍 Rust MCP Server + Inspector Example: SSE Transport with an Add Tool

What does this project do exactly?

It sets up an MCP server with an SSE transport, registers a tool to add two numbers, connects an MCP client, and prompts a RIG agent (using OpenAI) to invoke that tool.

What dependencies are required?

Rust with tokio, anyhow, serde_json, mcp_core, mcp_core_macros, and rig. The MCP Inspector requires Node.js to run npx.

What transport does the server use?

Server-Sent Events (SSE) on port 3001, with the inspector proxy listening on port 3000 and a web UI on port 5173.

Does this require an API key?

Yes, the RIG agent uses an OpenAI model, so a valid OpenAI API key must be set in the environment.

How can I test the tool manually?

After starting the server and inspector, visit http://localhost:5173 to see the tool list and invoke the Add tool. The agent also runs a prompt "Add 10 + 10" and prints the response.