Agents-OSS
EN
Submit
← Back to all projects

XcodeBuildMCP

Gives a coding agent the Xcode build, simulator and log commands it otherwise has to guess at

MCP serverMIT
Stars
6.3k
Forks
314
Open issues
22
Last commit
12 Aug 2026

What is XcodeBuildMCP?

A coding agent working on an iOS project can write Swift all day and never find out whether it compiles, because the commands that would tell it are long, particular and easy to get subtly wrong. This exposes them properly: build a scheme, boot a simulator, install and launch the app, read the logs, run the tests. Because the agent gets a result it can act on, the loop closes — it fixes the error it just caused instead of handing you a diff to try. It ships as both an MCP server and a plain command-line tool, with optional skills that tell the agent how to use whichever one you installed. It is macOS-only and tracks a current Xcode.

What can you do with XcodeBuildMCP?

  • Let the agent find out whether it compiled — Building a scheme returns real errors and warnings, so the agent corrects its own mistake in the same turn instead of handing you an untested diff.
  • Run the app, not just build it — A simulator can be booted and the app installed and launched, which is the difference between "it compiles" and "it starts".
  • Read the logs it just produced — Runtime logs come back to the agent, so a crash after launch is something it can investigate rather than something you have to relay.
  • Use it as a command or as a server — The same package is an MCP server for agents that speak MCP and a plain CLI for agents that are better at running commands.
  • Tell the agent how to drive it — Optional skills prime the agent with usage guidance, which matters for a tool surface this large — the tools exist, but knowing which to reach for is the hard part.

Before you choose XcodeBuildMCP

  • macOS only, and it tracks a current Xcode and a recent macOS — an older machine kept on a pinned toolchain for a legacy project is exactly the case it does not cover.
  • It gives an agent the ability to build and run code on your machine, so the usual command-approval settings in your client are the boundary — there is no sandbox in this layer.

Frequently asked questions

Is XcodeBuildMCP free for commercial use?

XcodeBuildMCP is released under the MIT licence — OSI-approved open source, which permits commercial use.

How can XcodeBuildMCP be deployed?

XcodeBuildMCP is available as Runs locally.

Documentation

Reproduced from the getsentry/XcodeBuildMCP README, published under MIT. Read the original ↗

A Model Context Protocol (MCP) server and CLI that provides tools for agent use when working on iOS and macOS projects.

CI pkg.pr.new

Installation

XcodeBuildMCP ships as a single package with two modes: a CLI for direct terminal use and an MCP server for AI coding agents. Either install method gives you both.

Option A — Homebrew

brew tap getsentry/xcodebuildmcp
brew install xcodebuildmcp

Option B — npm (Node.js 18+)

npm install -g xcodebuildmcp@latest

Verify either install:

xcodebuildmcp --help

Connect your MCP client

Drop-in config snippets for Cursor, Claude Code, Codex, can be found in the official docs page MCP Clients. Most clients can also run the MCP server on demand via npx -y xcodebuildmcp@latest mcp without a global install.

Requirements

  • macOS 14.5 or later
  • Xcode 16.x or later
  • Node.js 18.x or later (not required for Homebrew installation)

Skills

XcodeBuildMCP now includes two optional agent skills:

  • MCP Skill: Primes the agent with instructions on how to use the MCP server’s tools (optional when using the MCP server).

  • CLI Skill: Primes the agent with instructions on how to navigate the CLI (recommended when using the CLI).

To install with a global binary:

xcodebuildmcp init

Or install directly via npx without a global install:

npx -y xcodebuildmcp@latest init

For further information on installing skills, see Agent Skills.

Notes

  • XcodeBuildMCP requests xcodebuild to skip macro validation to avoid errors when building projects that use Swift Macros.
  • Device tools require code signing to be configured in Xcode. See Device Code Signing.

Privacy

XcodeBuildMCP uses Sentry for internal runtime error telemetry only. For details and opt-out instructions, see Privacy & Telemetry.

CLI

XcodeBuildMCP provides a unified command-line interface. The mcp subcommand starts the MCP server, while all other commands provide direct terminal access to tools:

# Install globally
npm install -g xcodebuildmcp@latest

# Start the MCP server (for MCP clients)
xcodebuildmcp mcp

# List available tools
xcodebuildmcp tools

# Build for simulator
xcodebuildmcp simulator build --scheme MyApp --project-path ./MyApp.xcodeproj

# Prepare portable test products without running tests
xcodebuildmcp simulator build --scheme MyApp --project-path ./MyApp.xcodeproj --simulator-name "iPhone 17" --build-for-testing --test-products-path ./MyApp.xctestproducts

# Run previously prepared test products
xcodebuildmcp simulator test --test-products-path ./MyApp.xctestproducts --simulator-name "iPhone 17"

Check for updates and upgrade in place:

xcodebuildmcp upgrade --check
xcodebuildmcp upgrade --yes

The CLI uses a per-workspace daemon for stateful operations (log capture, debugging, etc.) that auto-starts when needed. See the CLI guide for full documentation.

Documentation

XcodeBuildMCP
Ask AI
GitHub

Other options in Tools & Integrations

openclawTools & Integrations

OpenClaw installs a Gateway on your own machine and makes it the local control plane for sessions, tools, events and channel connections, so the Control UI, CLI, TUI and every connected messaging channel drive the same assistant. Models can be hosted or local, and the project expects new capabilities to arrive as plugins built on the plugin SDK rather than as changes to the core. The constraints are stated plainly in the README: it is designed for a single operator, tools run on the host for the main session unless you configure sandboxing, and DM-capable channels leave the assistant reachable by unknown senders until you approve a pairing with `openclaw pairing approve <channel> <code>`. The README tells you to read the security, exposure and sandboxing guides before connecting other users or exposing the Gateway remotely, which is a fair signal that shared or multi-user deployment is not the case it was built for.

OfficialMIT388k+893
hermes-agentTools & Integrations

Hermes Agent is Nous Research's own agent: a terminal UI plus a single gateway process that carries the same conversation into Telegram, Discord, Slack, WhatsApp, Signal and Email, with a cron scheduler, isolated subagents, and seven terminal backends from local and Docker to Modal and Vercel Sandbox. Its distinguishing piece is a closed learning loop — the agent curates its own memory on periodic nudges, creates skills after complex tasks, and searches past sessions through FTS5 with LLM summarization — which means the state that makes it useful accumulates under `~/.hermes`, outside your version control, and needs occasional pruning. It is an assistant you run and configure rather than a library you build on: the README documents `hermes` subcommands and slash commands, not an embedding API, so it is a poor fit if you wanted an agent loop to call from your own code. Putting a shell-capable agent behind chat platforms also makes the command-approval and DM-pairing settings load-bearing rather than optional.

OfficialMIT238k+3.8k
n8nWorkflow

Connects hundreds of services as workflows you draw on screen — each step in the flow is a node — and increasingly puts an AI agent node inside those workflows. The licence is the thing to check first: it is source-available under the Sustainable Use License, not OSI-approved open source. Internal use and self-hosting are fine; reselling it as a hosted service is not.

OfficialSource-available203k+1.2k