XcodeBuildMCP
by Sentry
XcodeBuildMCP is a Model Context Protocol server and CLI tool developed by Sentry (originally created by Cameron Cooke) that gives AI agents full control over Xcode for iOS and macOS development. With over 4,700 GitHub stars, it has become the go-to MCP server for Apple-platform agentic development. The server unlocks the complete developer loop directly from your AI coding assistant: build projects for simulators and real devices, run unit and UI tests, debug with LLDB breakpoints and variable inspection, take simulator screenshots, and deploy to physical devices over USB or Wi-Fi. XcodeBuildMCP operates as a per-workspace daemon that maintains state across interactions, enabling truly autonomous development workflows where your AI agent can iterate on compilation errors, fix failing tests, and verify UI changes without human intervention. It supports dual-mode operation as both an MCP server for AI clients like Claude Code, Cursor, VS Code, Windsurf, GitHub Copilot, and Xcode, and as a standalone CLI for terminal use. Project-level configuration via a YAML file allows you to specify schemes, project paths, and target simulators. The server handles code signing for device builds, supports Swift Macros, and includes Agent Skills that reduce upfront token cost by letting agents discover capabilities on demand. Requires macOS 14.5+, Xcode 16+, and Node.js 18+ (Node not needed for Homebrew installs).
Installation
Key Features
- ✓Build iOS and macOS projects for simulators and real devices with automatic error diagnosis
- ✓LLDB debugging with breakpoints, variable inspection, and step-through execution
- ✓UI automation including simulator screenshots and interface testing
- ✓Real device deployment over USB and Wi-Fi with code signing support
- ✓Dual-mode operation as MCP server for AI agents and standalone CLI for terminal use
- ✓Per-workspace daemon maintaining state across interactions for autonomous workflows
Use Cases
- →AI-assisted iOS and macOS app development with autonomous build-fix-test loops
- →Debugging Swift compilation errors and runtime crashes through AI agents
- →Running and verifying unit tests and UI tests from your AI coding assistant
- →Deploying builds to physical iOS devices for testing without leaving your editor
- →Automating repetitive Xcode workflows like scheme switching and simulator management