MCP cover image

用于模型上下文协议服务器和客户端的官方Swift SDK。与@loopwork-ai合作维护。

10

Github Watches

28

Github Forks

348

Github Stars

MCP Swift SDK

Swift implementation of the Model Context Protocol (MCP).

Requirements

  • Swift 6.0+ (Xcode 16+)

Platform Support

Platform Minimum Version
macOS 13.0+
iOS / Mac Catalyst 16.0+
watchOS 9.0+
tvOS 16.0+
visionOS 1.0+
Linux
Windows

[!IMPORTANT]
MCP's transport layer handles communication between clients and servers. The Swift SDK supports multiple transport mechanisms, with different platform availability:

  • StdioTransport is available on Apple platforms and Linux distributions with glibc, such as Ubuntu, Debian, Fedora, CentOS, or RHEL.

  • NetworkTransport is available only on Apple platforms.

Installation

Swift Package Manager

Add the following to your Package.swift file:

dependencies: [
    .package(url: "https://github.com/modelcontextprotocol/swift-sdk.git", from: "0.7.1")
]

Usage

Basic Client Setup

import MCP

// Initialize the client
let client = Client(name: "MyApp", version: "1.0.0")

// Create a transport and connect
let transport = StdioTransport()
try await client.connect(transport: transport)

// Initialize the connection
let result = try await client.initialize()

Basic Server Setup

import MCP

// Initialize the server with capabilities
let server = Server(
    name: "MyServer", 
    version: "1.0.0",
    capabilities: .init(
        prompts: .init(),
        resources: .init(
            subscribe: true
        ),
        tools: .init()
    )
)

// Create transport and start server
let transport = StdioTransport()
try await server.start(transport: transport)

// Register method handlers
server.withMethodHandler(ReadResource.self) { params in
    // Handle resource read request
    let uri = params.uri
    let content = [Resource.Content.text("Example content")]
    return .init(contents: content)
}

// Register notification handlers
server.onNotification(ResourceUpdatedNotification.self) { message in
    // Handle resource update notification
}

// Stop the server when done
await server.stop()

Working with Tools

// List available tools
let tools = try await client.listTools()

// Call a tool
let (content, isError) = try await client.callTool(
    name: "example-tool", 
    arguments: ["key": "value"]
)

// Handle tool content
for item in content {
    switch item {
    case .text(let text):
        print(text)
    case .image(let data, let mimeType, let metadata):
        // Handle image data
    }
}

Working with Resources

// List available resources
let (resources, nextCursor) = try await client.listResources()

// Read a resource
let contents = try await client.readResource(uri: "resource://example")

// Subscribe to resource updates
try await client.subscribeToResource(uri: "resource://example")

// Handle resource updates
await client.onNotification(ResourceUpdatedNotification.self) { message in
    let uri = message.params.uri
    let content = message.params.content
    // Handle the update
}

Working with Prompts

// List available prompts
let (prompts, nextCursor) = try await client.listPrompts()

// Get a prompt with arguments
let (description, messages) = try await client.getPrompt(
    name: "example-prompt",
    arguments: ["key": "value"]
)

Changelog

This project follows Semantic Versioning. For pre-1.0 releases, minor version increments (0.X.0) may contain breaking changes.

For details about changes in each release, see the GitHub Releases page.

License

This project is licensed under the MIT License.

相关推荐

  • https://maiplestudio.com
  • Find Exhibitors, Speakers and more

  • Joshua Armstrong
  • Confidential guide on numerology and astrology, based of GG33 Public information

  • Alexandru Strujac
  • Efficient thumbnail creator for YouTube videos

  • Elijah Ng Shi Yi
  • Advanced software engineer GPT that excels through nailing the basics.

  • Emmet Halm
  • Converts Figma frames into front-end code for various mobile frameworks.

  • lumpenspace
  • Take an adjectivised noun, and create images making it progressively more adjective!

  • https://zenepic.net
  • Embark on a thrilling diplomatic quest across a galaxy on the brink of war. Navigate complex politics and alien cultures to forge peace and avert catastrophe in this immersive interstellar adventure.

  • https://reddgr.com
  • Delivers concise Python code and interprets non-English comments

  • Beria Joey
  • 你的职业规划师,不走弯路就问我。Sponsor:小红书“ ItsJoe就出行 ”

  • Mintplex-Labs
  • 带有内置抹布,AI代理,无代理构建器,MCP兼容性等的多合一桌面和Docker AI应用程序。

  • ravitemer
  • 一个功能强大的Neovim插件,用于管理MCP(模型上下文协议)服务器

  • jae-jae
  • MCP服务器使用剧作《无头浏览器》获取网页内容。

  • patruff
  • Ollama和MCP服务器之间的桥梁,使本地LLMS可以使用模型上下文协议工具

  • pontusab
  • 光标与风浪冲浪社区,查找规则和MCP

  • WangRongsheng
  • 🧑‍🚀 llm 资料总结(数据处理、模型训练、模型部署、 o1 模型、mcp 、小语言模型、视觉语言模型)|摘要世界上最好的LLM资源。

  • n8n-io
  • 具有本机AI功能的公平代码工作流程自动化平台。将视觉构建与自定义代码,自宿主或云相结合,400+集成。

  • av
  • 毫不费力地使用一个命令运行LLM后端,API,前端和服务。

  • metorial
  • 数百个MCP服务器的容器化版本📡📡

  • appcypher
  • 很棒的MCP服务器 - 模型上下文协议服务器的策划列表

    Reviews

    2 (1)
    Avatar
    user_93NPwRkt
    2025-04-17

    I've been using the Swift-SDK by ModelContextProtocol and I'm thoroughly impressed. The integration process was seamless, and the SDK provides robust features for swift application development. The documentation on their GitHub page is superb, making implementation straightforward even for new users. Highly recommend checking it out for any Swift projects!