apple-appstore-toolkit

Apple App Store Toolkit

A Claude Code plugin that reviews iOS apps for Apple App Store readiness using 8 specialized review agents. Supports both React Native and Swift/Xcode projects.

Install

In Claude Code, run:

/plugin marketplace add crgeee/apple-appstore-toolkit
/plugin install apple-appstore-toolkit

Then navigate to your iOS project and run:

/apple-appstore-toolkit:review-app

git clone https://github.com/crgeee/apple-appstore-toolkit.git
claude --plugin-dir /path/to/apple-appstore-toolkit

Prerequisites

• Claude Code CLI installed
• An iOS project (React Native or Swift/Xcode) in your working directory

Why?

Apple rejects approximately 25% of all App Store submissions (1.93 million out of 7.77 million in 2024). The top causes — missing privacy manifests, incorrect Info.plist keys, absent restore-purchases buttons — are all detectable before you submit.

This toolkit catches those issues during development so you don't waste days waiting for a rejection email.

How It Works

/review-app command
       │
       ├── Detects project type (React Native vs Swift/Xcode)
       │
       ├── Launches specialized agents (parallel by default)
       │   ├── info-plist-analyzer
       │   ├── privacy-compliance-reviewer
       │   ├── ui-ux-guidelines-reviewer
       │   ├── performance-stability-reviewer
       │   ├── assets-metadata-reviewer
       │   ├── iap-compliance-reviewer
       │   ├── security-reviewer
       │   └── react-native-reviewer  (only for RN projects)
       │
       └── Aggregates findings into a unified report
           ├── Critical Issues (will cause rejection)
           ├── Important Issues (likely rejection)
           ├── Advisory (best practices)
           └── Passed Checks

Each agent runs independently with its own scope, then results are combined by severity. Every issue includes the file location, a concrete fix suggestion, and the Apple guideline reference.

Agents

Note: The privacy-compliance-reviewer is configured to use the Opus model family (rather than inheriting your session model) for higher accuracy on complex privacy manifest analysis. All other agents inherit the model you're running Claude Code with.

Usage

Full Review (default — parallel)

/apple-appstore-toolkit:review-app

Sequential Review

/apple-appstore-toolkit:review-app sequential

Targeted Review

/apple-appstore-toolkit:review-app privacy security    # specific agents
/apple-appstore-toolkit:review-app plist               # single agent
/apple-appstore-toolkit:review-app reactnative         # RN-specific checks
/apple-appstore-toolkit:review-app iap assets          # multiple agents

Available Aspects

Example Output

# App Store Readiness Report

**Project Type:** React Native (Expo)
**Agents Run:** 8 of 8
**Overall Assessment:** Needs Work

## Critical Issues (3 found)