A GitHub Script Catches App Store and Google Play Rejection Reasons Before Review: 542 Stars in a Day
An App Store rejection costs you a week. You submit, wait for review, get a one-line rejection with no explanation, and start guessing. The mjmirza/app-store-compliance repository, which gained 542 stars in a day, offers a bash script that scans your project before submission and lists critical risks with the file and line. It checks two things first: the demo account, and whether the privacy declaration matches what the app actually does. A hook for AI agents such as Claude Code halts the upload until the risk is resolved.
AI-processed from GitHub Trending AI; edited by Hamidun News
Building a mobile app in an evening with Claude Code or Cursor is realistic today. Getting through App Store or Google Play review on the first try is not always. The mjmirza/app-store-compliance repository, which gained 542 stars on GitHub Trending in a single day, tries to close exactly that gap. It is an open collection of rules plus a script that checks your project before submission, not after a reviewer has written a rejection.
What problem it solves
Anyone who has shipped an app knows the typical cycle. You submit a build and wait a few days. A one-line rejection arrives, citing a guideline section. You guess what exactly is wrong, make fixes at random, resubmit, and wait again. The repository's author estimates one such loop at about a week.
The author cites figures given in the README. Over one year, Apple reviewed about 7.77 million submissions and rejected roughly 1.93 million, almost one in four. In 2025, Google blocked more than 1.75 million submissions for policy violations. According to the same data, the number of releases in the first quarter of 2026 grew by about 60% year over year, and on iOS by almost 80%. The author ties the surge to AI coding tools and claims the stores have tightened their rules rather than loosened them. These figures come from the repository itself, and we did not check them against primary sources.
How it works
Installation takes two commands. The first clones the repository; the second runs the app-store-compliance-guard.sh script on the folder containing your Xcode project or Android app. Only git and bash are required. The script installs nothing and changes nothing in the project. It checks a small project in about a second and a large one in about a minute. Findings are listed in descending order of severity, critical ones first, each with a file and line. At the end, the script reports whether the app is safe to submit.
Rejection patterns are collected in the rejection-patterns.json file. Each one is tied to a specific rule and a suggested fix. That turns a one-line rejection into a known problem with a known solution. The script prioritizes two causes that the author considers the most common:
- a demo account that works against a live backend;
- a privacy declaration that matches what the app actually does, including every SDK it integrates.
Among other things, the README mentions a missing privacy manifest, which Apple has required since 2024. Another rule: since late 2025, an app is rejected if it sends personal data to an external AI without a consent prompt that names the provider. For anyone embedding an LLM in a mobile product, this is an easy requirement to miss.
A hook for AI agents
The most practical feature of the repository is that the script can be installed as a hook. The install.sh installer supports Claude Code, Codex, Gemini CLI, OpenCode, Cursor, GitHub Copilot, Grok, Qwen Code, Kimi Code, Factory Droid, Auggie and Windsurf.
The hook fires before submission commands such as fastlane, eas submit, gradle bundleRelease and xcrun altool. If it finds a critical risk, the agent will not run the upload and will show the fix instead. For Kiro, Cline and several other tools there is no auto-install; you have to wire it up manually or run the script before submitting.
After installation, the installer checks on its own that the hook blocks a test submission command. There is also a one-off mode that needs no installation: paste a ready-made prompt into your agent, and it runs a read-only audit.
Limitations
The scanner sees code but not account state. It cannot tell whether your developer account is active, whether all agreements have been accepted, or whether package names are registered in Play, any of which can delay a first launch by weeks. For that there is a separate checklist, PRE-SUBMISSION-CHECKLIST.md, which you have to go through manually.
Anything the tool cannot verify, it flags as needing review rather than treating it as clean. The authors left this behavior in deliberately: an extra warning costs a minute, while a missed one costs a rejection. The README includes reviews of first-time approvals, one from Reddit and one in German from a chat, but these remain isolated anecdotes rather than a measurement of effectiveness. One person maintains the project, so how current the rules stay depends on his updates. There is a monitor.py script for that, which checks the project against Apple Developer's RSS feed, but its quality cannot be judged from the README.
What to do
If you are preparing a release, run the script on your project before submitting. It takes about a minute and breaks nothing. If an agent built the app, install the hook so the check runs automatically. Separately, check the two things that, in the author's view, trip up most developers: a working demo account for the reviewer, and a privacy declaration that matches the SDKs actually in use, especially if the app sends data to an LLM.
- Nov 19, 2026Amazon and Google's Anthropic Stakes Valued at $308 Billion in IPO at $83 a Share
- Nov 19, 2026Google's Gemini 4 Neon runs twice as fast as Carbon and costs 17% less per task
- Nov 19, 2026Nvidia posts record $115.8 billion revenue, yet trades cheaper than at any point since 2012
- Nov 19, 2026After an Anthropic ban, Codex moved 43 Claude Code sessions in 20 minutes
Key Questions Answered
What problem it solves?
Anyone who has shipped an app knows the typical cycle. You submit a build and wait a few days. A one-line rejection arrives, citing a guideline section. You guess what exactly is wrong, make fixes at random, resubmit, and wait again. The repository's author estimates one such loop at about a week. The author cites figures given in the README. Over one year, Apple reviewed about 7.77 million submissions and rejected roughly 1.93 million, almost one in four.
How it works?
Installation takes two commands. The first clones the repository; the second runs the app-store-compliance-guard.sh script on the folder containing your Xcode project or Android app. Only git and bash are required. The script installs nothing and changes nothing in the project. It checks a small project in about a second and a large one in about a minute. Findings are listed in descending order of severity, critical ones first, each with a file and line.
What are the key details on A hook for AI agents?
The most practical feature of the repository is that the script can be installed as a hook. The install.sh installer supports Claude Code, Codex, Gemini CLI, OpenCode, Cursor, GitHub Copilot, Grok, Qwen Code, Kimi Code, Factory Droid, Auggie and Windsurf. The hook fires before submission commands such as fastlane, eas submit, gradle bundleRelease and xcrun altool. If it finds a critical risk, the agent will not run the upload and will show the fix instead.
Build & Ship Production AI Agents Without Boilerplate Code
Start your free 72-hour trial at Hamidun Academy (academy.hamidun.com) — interactive lessons on Claude Code & Cursor, live Tech Radar, and production agent templates.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.