AI Tools for Mobile Development
CLI tools, MCP servers, and agent skills to supercharge your mobile workflow.
11 tools curatedXcodeBuildMCP
Lets your AI agent trigger Xcode builds, run tests, and read diagnostics directly. Useful when you want Claude or Cursor to fix a build error without copy-pasting logs.
Mobile MCP
An MCP server for mobile automation and scraping across iOS and Android. Works with emulators, simulators, and real devices.
App Store Connect CLI (asc)
Manage TestFlight, builds, certificates, and metadata from the terminal instead of clicking through App Store Connect in a browser.
SwiftUI Agent Skill
Teaches your AI agent how SwiftUI actually works, including Liquid Glass patterns and state management. The kind of context that keeps it from suggesting UIKit workarounds.
Expo Skills
Maintained by the Expo team. Gives AI agents the context they need to scaffold, build, and debug Expo apps without hallucinating deprecated APIs.
Swift Concurrency Agent Skill
Gives your AI agent deep knowledge of async/await, actors, and Sendable. Especially useful during Swift 6 migrations where the compiler errors are not exactly self-explanatory.
Android MCP Server
Gives AI agents access to a connected Android device through ADB. They can take screenshots, inspect the UI hierarchy, and manage packages on your behalf. Supports multiple devices at once.
iOS Simulator Skill
Your AI agent can boot simulators, navigate by accessibility labels, run visual diffs between screenshots, and trigger builds. Good for automating the test-fix-test cycle.
DroidMind
Connects AI agents to Android devices over MCP for things like tapping through UI flows, pulling logs, and running shell commands. Think of it as giving your agent a phone to play with.
MCP Flutter
Pairs an MCP server with a Flutter toolkit so your AI agent can grab screenshots, read widget trees, and catch runtime errors from a running Flutter app.
Claude Swift Engineering
Loads Claude Code with agents and knowledge for Swift 6.2, TCA, SwiftUI, and concurrency. Covers the whole loop from architecture planning to testing so the agent doesn't wing it.