SuperAPI AI Directory

xcode-mcp-server

Xcode integration for project management, file operations and build automation.

Open website

About this tool

Purpose: xcode-mcp-server is a server based on the Model Context Protocol (MCP) that enables AI assistants to interact with Xcode projects, completing tasks such as project management, file operations, build testing, dependency management, and simulator control.

Main capabilities: In project management, it can set the active project, read project information, create iOS/macOS/watchOS/tvOS projects from templates, add files to a project with specified target and group, resolve workspace-associated projects, and list schemes. File operations support reading and writing with different encodings, base64 handling of binary files, searching text by pattern or regex, checking file existence and metadata, and automatically creating directories. Build and test support custom builds, running tests and reporting failures, code analysis, cleaning build directories, and archiving for distribution. For dependencies, it supports CocoaPods initialization, installing updates, adding and removing pods, and executing pod commands; Swift Package Manager supports initializing packages, adding and removing dependencies, updating resolution, generating documentation with DocC, and running tests and builds. Simulator tools can list, launch, and shut down simulators, install and launch apps, take screenshots and record video, and manage settings and state. There are also tools for executing xcrun, compiling asset catalogs, generating icon sets, performance tracing, exporting and validating archives, and switching Xcode versions.

Usage: Requires macOS with Xcode 14.0+, Node.js 16+; Swift Package Manager features require Swift 5.5+; CocoaPods is optional. The included setup script can automatically complete environment checks, dependency installation, configuration, and Claude Desktop integration; you can also manually run git clone https://github.com/r-huijts/xcode-mcp-server.git、cd xcode-mcp-server, npm install, npm run build, and create a .env to configure parameters such as PROJECTS_BASE_DIR. After startup, the AI assistant connects through the server URL (usually http://localhost:3000)连接。

Limitations: Only supports macOS and the Xcode environment; file access is restricted by allowed directories; the specific tool list and usage must be checked in the repository documentation, and the materials do not provide a complete tool list or all parameter details.