XcodeBuildMCP

XcodeBuildMCP is an open-source toolkit designed to enable AI agents to automate the full range of Xcode workflows for iOS and macOS app development. By providing both an MCP server and a command-line interface, it allows AI assistants to independently build, test, debug, and deploy applications without human intervention. The platform is intended for developers seeking to integrate AI-driven automation into their Xcode-based projects, streamlining the development lifecycle from project creation to deployment on simulators and physical devices.

Key features include deep interoperability with Xcode, allowing two-way integration with Xcode's coding agents and development tools. 3's Claude and Codex agents, automatically detecting the active scheme and simulator to facilitate operations. It supports LLDB debugging, enabling agents to attach the debugger, set breakpoints, inspect variables, and execute LLDB commands. The toolkit also provides UI automation capabilities, such as interacting with simulator UI elements, capturing screenshots, and automating user interface testing workflows. Real device testing is supported, allowing deployment and testing on physical devices over USB or Wi-Fi, with comprehensive log capture and debugging.

The platform offers multi-client support, working with tools such as Cursor, Claude Code, VS Code, Windsurf, Xcode, GitHub Copilot, Codex, Amp, and OpenCode, as well as any MCP-compatible client. XcodeBuildMCP can be installed as an MCP server for AI coding agents or globally for direct CLI usage, and it is suitable for scripting and CI/CD pipelines. Configuration is handled at the project level using a YAML file, enabling agents to manage schemes, simulators, and workflows efficiently. The tool also includes features like selective workflow loading and agent skill files to optimize AI agent performance.

XcodeBuildMCP is distributed under the MIT license and is actively developed as an open-source, community-driven project. It is available for installation via NPX or Homebrew, and contributions from the community are encouraged. The platform aims to bridge the gap between AI assistants and Xcode, supporting autonomous development workflows for iOS and macOS applications.