Open-Source Testing · conorluddy

ios-simulator-skill

iOS Simulator Skill is a Python-based toolset for Claude Code that automates iOS app building, testing, and simulator interaction via xcodebuild and xcrun simctl. It provides 27 scripts for build automation, device state management, UI navigation, testing, and permissions handling, optimized for token efficiency in AI agent workflows.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/conorluddy/ios-simulator-skill

1.1k

GitHub stars

Forks

Python

Primary language

MIT

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	conorluddy/ios-simulator-skill
Owner	conorluddy
Primary language	Python
License	MIT — OSI-approved
Stars	1.1k
Forks	75
Open issues	19
Latest release	v1.4.0 (2026-04-12)
Last updated	2026-06-18
Source	https://github.com/conorluddy/ios-simulator-skill

What ios-simulator-skill is

A Claude Code skill wrapping macOS xcodebuild, xcrun simctl, and idb tools into 27 Python scripts with progressive error disclosure, accessibility-based UI navigation, and token-optimized output. Supports Xcode project builds, test parsing, simulator lifecycle, gesture automation, and app sandbox inspection with tunable environment variables for CI/local development tradeoffs.

Quickstart

Get the ios-simulator-skill source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/conorluddy/ios-simulator-skill.git2cd ios-simulator-skill3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

AI-assisted iOS development in Claude Code

Enables Claude to autonomously build, test, and interact with iOS simulators without context-heavy raw tool output. Accessibility-first navigation and progressive error disclosure reduce token consumption by 96–99% vs. raw xcodebuild and screenshot output.

Automated iOS CI/testing pipelines on macOS runners

Provides structured, tunable wrappers around xcodebuild, simctl, and test result parsing for GitHub Actions, GitLab CI, or local development. Environment variables allow CI tuning (boot timeouts, output caps) independent of local defaults.

Accessibility and permissions testing at scale

Semantic UI navigation via accessibility APIs, WCAG compliance auditing, and fine-grained permission grant/revoke workflows support QA automation and accessibility validation without brittle pixel-coordinate tapping.

Implementation considerations

Requires macOS 12+ with Xcode Command Line Tools pre-installed; IDB (optional, for interactive features) via Homebrew; Python 3 and Pillow (optional, for visual diffs) via pip.
Installation via Claude Code plugin marketplace (recommended) or manual git clone into ~/.claude/skills/ or project .claude/skills/. Restart Claude Code to auto-load.
All operational limits (boot timeouts, output caps, polling intervals, cache size) tunable via IOS_SIM_* environment variables; defaults optimized for Apple Silicon local development; CI runners or large monorepos may need tuning.
Token efficiency depends on script choices: accessibility-based navigation returns ~10 tokens vs. 1,600–6,300 for raw screenshots. Progressive error disclosure (build summary, then drill-in on demand) keeps agent context tight.
27 scripts cover xcodebuild, simulator lifecycle, UI automation, testing, permissions, and debugging; --json and --help flags on all scripts for machine-readable output and discoverability.

When to avoid it — and what to weigh

Non-macOS development environments — Requires macOS 12+, Xcode Command Line Tools, and xcrun simctl. Not applicable to Linux, Windows, or cloud-only CI runners without macOS agents.
Heavy reliance on visual pixel-perfect testing — While visual_diff.py exists, the skill prioritizes accessibility-based navigation and token optimization. If your app's correctness depends on exact pixel layouts (design-focused QA), raw screenshot tools may be more direct.
Projects requiring real device testing — Targets iOS Simulator only (via simctl and idb). Real device deployment, TestFlight workflows, and physical device interaction are out of scope.
Offline or air-gapped CI environments — Requires macOS agent with Xcode tooling, internet for plugin marketplace installation, and optional external tools (idb, Pillow). Pre-cached or vendored setup may be complex.

License & commercial use

MIT License (MIT). Permissive OSI-compliant open-source license allowing commercial use, modification, and distribution with attribution and no warranty.

MIT License is a permissive OSI license explicitly allowing commercial use in proprietary products and services, provided the MIT license text and copyright notice are retained. No restrictions on profit-driven deployment. However, review project-specific dependencies (idb, Pillow) for their own license compatibility in your commercial context.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Strong
License clarity	Clear
Deployment complexity	Low
DEV.co fit	Good
Assessment confidence	High

Security considerations

Executes xcodebuild, xcrun simctl, and optional idb commands on the host macOS system with user privileges. No explicit security audit or vulnerability disclosure process documented. Simulator automation inherently runs untrusted app code in an emulated environment; isolation is a property of iOS Simulator, not the skill itself. For production CI, apply standard macOS hardening (code signing, access controls, audit logging) and review the skill code before granting shell access to sensitive build systems. Accessibility tree inspection (navigator.py) is safe; idb depends on Facebook's security posture.

Alternatives to consider

xclaude-plugin (by same author)

Lighter-weight Xcode build tooling plugin without simulator scripts; if you only need xcodebuild wrapping and progressive error disclosure, this is a leaner alternative.

XC-MCP (by same author)

Model Context Protocol version of iOS Simulator tooling; use if you prefer MCP-based integration over Claude Code skills, or need cross-model compatibility.

Native Xcode UI automation (XCUITest) + shell scripting

Lower-level, no external dependency on Claude Code or AI agent framework; suitable if you control test code directly and prefer standard iOS testing paradigms over AI-agent-optimized workflows.

Software development agency

Build on ios-simulator-skill with DEV.co software developers

Install iOS Simulator Skill in Claude Code to automate builds, tests, and simulator interaction with 96–99% token savings. Perfect for AI-assisted workflows and macOS CI pipelines.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

app-store-connect-cli-skills claude-code-plugins-plus-skills playwright-skill

Related on DEV.co

Explore the category and the services that help you build with it.

Open-Source Testing AI Development Application Development DevOps & CI/CD Talk to DEV.co

ios-simulator-skill FAQ

Can this work outside Claude Code?

The skill is designed for Claude Code. The individual Python and shell scripts are callable from any CLI, but the 'skill' integration is Claude Code–specific. XC-MCP is the author's MCP alternative for other AI contexts.

Do I need idb and Pillow installed?

No. idb (interactive features) and Pillow (visual diffs) are optional. Core build, simctl, and accessibility-based navigation work with just xcodebuild, xcrun simctl, and Python 3.

How much does this reduce token usage compared to raw tools?

96–99% reduction in typical workflows: screen analysis 97.5%, find-and-tap 99%, login flow 96%. Achieved via progressive error disclosure, accessibility tree output instead of screenshots, and output resizing/compression.

What if my simulator is slow on CI?

Tune IOS_SIM_BOOT_TIMEOUT (default 300s) higher for slow runners; GitHub-hosted macOS can need 4–6 min cold-start. Other timeouts (ERASE, SUBPROCESS) are tunable per operation.

Software development & web development with DEV.co

DEV.co is a software development agency delivering custom software development services to companies building on open source. Our software developers and web developers design, integrate, and ship production systems — spanning web development, APIs, AI, data, and cloud. If ios-simulator-skill is part of your open-source testing roadmap, our team can implement, customize, migrate, and maintain it.

Streamline iOS Development with AI Agents

Install iOS Simulator Skill in Claude Code to automate builds, tests, and simulator interaction with 96–99% token savings. Perfect for AI-assisted workflows and macOS CI pipelines.

Talk to DEV.co Schedule a Consultation