A release with android:debuggable="true" lets anyone attach a debugger to your app, and an exported content provider hands its data to every other app on the phone. This skill finds those and nine other manifest mistakes before you publish to Google Play, and tells you how to fix each one.

Your agent runs a scanner on your Android project: every AndroidManifest.xml (native Android, Flutter, React Native, library modules) and every network security config under res/xml. It reads XML with Python's own parser, so comments, attribute order and tags over several lines do not matter, and it shows the line of the attribute that fired. Debug and test source sets are read but not reported. It never runs Gradle or the app. It returns BLOCK, REVIEW, or NOT FLAGGED, with the file, line, reason, and a safer approach. Secrets are never printed.

What it checks

BLOCK (fix before shipping)

• M01: android:debuggable="true"

• M02: cleartext HTTP allowed (usesCleartextTraffic or cleartextTrafficPermitted)

• M03: a content provider exported with no permission

• M04: user-installed certificates trusted outside debug-overrides

REVIEW (check before release)

• M11: app data copied to backups (allowBackup with no backup rules)

• M12: a service, receiver or activity exported with no permission

• M13: an intent-filter with no android:exported (fails on Android 12+)

• M14: android:testOnly="true"

• M15: Google Play restricted permissions (SMS, call log, QUERY_ALL_PACKAGES, all files access)

• M16: an https deep link without android:autoVerify

• M17: a secret in meta-data

What's included

• The scanner (Python 3.9+, standard library only)

• A rules guide with the reason and safer approach for each rule

• A risky and a safe example app manifest and network security config, and the real report the scanner produces on the risky ones

Limits

It checks each source manifest on its own, not the merged manifest, so attributes added by libraries, build types and Gradle placeholders are not seen. Java and Kotlin code is not checked. NOT FLAGGED is not a guarantee.

Contents

One ZIP file with 10 files: SKILL.md, the scanner in scripts/ (2 Python files), references/rules.md (every rule with its reason and safer approach), and examples/ (5 example files plus report.md, the real scanner output on the risky example).

How to use: This is an agent skill: a folder with a SKILL.md file and Python scripts that an AI coding agent such as Claude Code, Cursor, or Codex CLI can run. Unzip it into your agent's skills folder (for Claude Code: ~/.claude/skills/), then ask your agent to check your project. The scripts also run on their own with Python 3.9 or newer, with nothing to install. They read your files only: no network access, and nothing is sent anywhere.