conorluddy/ios-simulator-skill

An IOS Simulator Skill for ClaudeCode. Use it to optimise Claude's ability to build, run and interact with your apps, and to proxy xcodebuild to save token and context wastage.

What it solves

This project provides a set of automation tools designed to help AI agents (specifically Claude Code) and human developers build, test, and interact with iOS applications within the Xcode Simulator. It solves the problem of "token bloat" and fragility in AI-driven mobile development by replacing raw, verbose tool output and coordinate-based tapping with optimized, semantic summaries and accessibility-based navigation.

How it works

The skill consists of 29 Python and shell scripts that wrap standard Apple tools like xcodebuild and simctl, as well as the idb (iOS Debug Bridge) library.

  • Progressive Disclosure: Instead of dumping massive build logs, it provides a concise summary and allows the user or agent to request specific errors or warnings using a unique ID.
  • Semantic Navigation: It uses iOS accessibility APIs to find UI elements by their meaning (e.g., finding a button labeled "Login") rather than relying on unstable screen coordinates.
  • Token Optimization: It automatically resizes and compresses screenshots and truncates long logs to reduce the amount of data sent to the LLM, significantly lowering token costs.
  • Environment Control: It allows fine-tuning of timeouts and output caps via environment variables to adapt to different hardware (e.g., Apple Silicon vs. slow CI runners).

Who it’s for

It is primarily built for developers using Claude Code to automate iOS app development, as well as QA engineers and developers who want a more robust, scriptable way to manage the iOS Simulator and Xcode builds.

Highlights

  • 29 specialized scripts covering the full lifecycle: build, device state, navigation, testing, and lifecycle management.
  • Accessibility-first interaction for robust UI automation that doesn't break when the layout changes.
  • Token-efficient output that reduces raw tool output by up to 99% for common tasks.
  • Comprehensive device control, including GPS simulation, dark mode switching, push notification triggering, and permission management.
  • Integrated health checks to verify the installation of Xcode, Python, and idb.

Related

  • Project
    MoonshotAI/FlashKDA

    A high-performance CUDA kernel implementation of Kimi Delta Attention (KDA) built on CUTLASS for NVIDIA SM90+ GPUs, serving as an optimized backend for the flash-linear-attention library.

  • Project
    zinggAI/zingg

    An ML-based tool for master data management and entity resolution that deduplicates and links records across disparate data sources at scale.

  • Project
    ModelCloud/GPTQModel

    A comprehensive LLM quantization toolkit that compresses models using methods like GPTQ and AWQ with hardware acceleration for NVIDIA, AMD, Intel, and Huawei NPUs.

  • Project
    WhiteNightShadow/hello_js_reverse_skill

    A comprehensive reverse engineering toolkit for Web interfaces that helps analyze obfuscated JS and restore API signatures using an anti-detection browser workflow.