About MCP-Go SDK
MCP-Go SDK is a Go toolkit for developers building MCP tools and servers. It defines tool interfaces and response formats, and includes a transport layer with stdio support.
Key features
- 01Build MCP tools and servers
- 02Define tool schemas and execution
- 03Format MCP tool responses
- 04Use a transport interface
- 05Included stdio transport
Use cases
- 01Implement an MCP tool in Go
- 02Create a minimal example server
- 03Expose tool parameters through JSON schemas
- 04Connect a command-line tool over stdio
MCP support
The SDK provides building blocks for implementing MCP-compliant tools and servers. Its transport layer includes stdio support for command-line tools.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEMCP-Go SDK
A Go SDK for building Model Communication Protocol (MCP) tools and servers. This SDK provides the building blocks for implementing MCP-compliant tools that can be used with AI applications like Cursor IDE.
Quick Start
Check out the example server in servers/example to see a minimal implementation of an MCP tool that echoes back messages:
package main
import (
"encoding/json"
"mcp-go/server"
"mcp-go/transport"
)
// EchoTool implements a simple echo tool
type EchoTool struct{}
func (t *EchoTool) Name() string {
return "echo"
}
func (t *EchoTool) Description() string {
return "A simple echo tool that returns the input message"
}
func (t *EchoTool) Schema() json.RawMessage {
return json.RawMessage(`{
"type": "object",
"properties": {
"message": {
"type": "string",
"description": "The message to echo back"
}
},
"required": ["message"]
}`)
}
func (t *EchoTool) Execute(params json.RawMessage) (interface{}, error) {
var input struct {
Message string `json:"message"`
}
if err := json.Unmarshal(params, &input); err != nil {
return nil, err
}
return map[string]interface{}{
"content": []map[string]interface{}{
{
"type": "text",
"text": input.Message,
},
},
"metadata": map[string]interface{}{
"length": len(input.Message),
},
}, nil
}
func main() {
// Create a new server with stdin/stdout transport
srv := server.NewServer(transport.NewStdioTransport())
// Register your tool
if err := srv.RegisterTool(&EchoTool{}); err != nil {
panic(err)
}
// Start the server
if err := srv.Start(); err != nil {
panic(err)
}
}
## Core Concepts
### 1. Tools
A Tool in MCP is a service that can be called by AI applications. Each tool must implement the `mcp.Tool` interface:
```go
type Tool interface {
// Name returns the unique identifier for this tool
Name() string
// Description returns a human-readable description
Description() string
// Schema returns the JSON schema for the tool's parameters
Schema() json.RawMessage
// Execute runs the tool with the given parameters
Execute(params json.RawMessage) (interface{}, error)
}
2. Response Format
Tools should return responses in the MCP format:
{
"content": [
{
"type": "text",
"text": "Your response text"
}
],
"metadata": {
// Optional metadata about the response
}
}
3. Transport Layer
The SDK provides a flexible transport layer through the Transport interface:
type Transport interface {
Send(data interface{}) error
Receive() ([]byte, error)
Close() error
}
By default, the SDK includes a stdio transport (transport.NewStdioTransport()) for command-line tools.
4. Configuration
To use your MCP tool with Cursor IDE, create a .cursor/mcp.json in your project root:
{
"mcpServers": {
"mytool": {
"command": "mytool",
"args": [],
"env": {}
}
}
}
Contributing
- Fork the repository
- Create a feature branch
- Add tests for new functionality
- Submit a pull request
Please ensure your code:
- Follows Go best practices
- Includes appropriate documentation
- Has test coverage
- Handles errors appropriately
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.