ios_skill

Overview

This skill provides iOS platform-specific development automation using XcodeBuildMCP tools for simulator, device, UI automation, and debugging. It streamlines building, testing, deploying, and automating UI interactions for iPhone apps. Use it to speed iterative development cycles and enforce consistent device validation before release.

How this skill works

Before any implementation, create a team of three agents (explore-codebase, research-expert, and mcp__XcodeBuildMCP__discover_projs) to analyze the codebase, verify iOS docs, and locate Xcode projects. The skill exposes MCP commands for simulator workflows (build_sim, boot_sim, launch_app_sim, test_sim), device operations (build_device, install_app_device, list_devices), and UI automation (tap, swipe, screenshot, snapshot_ui). After changes, run a sniper validation agent to verify integration and tests.

When to use it

• Building and compiling iPhone apps for simulator and device
• Running automated test suites on simulator or physical hardware
• Deploying and validating builds on connected iOS devices
• Automating UI flows for regression testing or demos
• Debugging app behavior and capturing view hierarchies or screenshots

Best practices

• Run an initial TeamCreate workflow to spawn explore, research, and discovery agents before coding
• Always validate a local build before committing changes
• Use simulator for fast iteration, then test on real devices prior to release
• Add stable accessibility IDs to support reliable UI automation
• Capture screenshots and view hierarchies to document UI states and debug regressions

Example use cases

• Build and run unit and UI tests on the iOS simulator as part of CI validation
• Install a nightly build to a QA device farm and run automated tap/swipe scenarios
• Capture screenshots of core flows to include in release notes or bug reports
• Use snapshot_ui to extract the view hierarchy when debugging layout or accessibility issues
• Run LLDB-assisted debugging on a device after reproducing a crash with automated steps