| name | android-play-production-checklist |
| description | Run a staged Google Play production-readiness checklist for Expo or React Native Android releases and launch blockers. |
Android Play Production Checklist
Use this as the top-level readiness orchestrator for releasing an Expo / React Native app to Google Play.
The purpose is not to recite every possible Google Play rule. The purpose is to produce a useful release verdict, identify the highest-risk blockers, and progressively disclose only the checklist areas that matter for this app.
Core Principle
A Google Play release is ready only when five things are true:
- the binary is technically valid and signed correctly
- the Play Console submission packet is complete
- policy declarations match actual app behavior
- reviewers and testers can access the full app
- production rollout risk is understood and controlled
Do not let successful eas build or successful upload be mistaken for production readiness.
Progressive Disclosure Model
Level 0 — Verdict
Start with one of these verdicts:
| Verdict | Meaning |
|---|
Ready to submit | No known blockers; remaining items are routine checks. |
Probably ready, verify these items | No obvious blockers, but important evidence is missing. |
Not ready | One or more blockers would likely prevent review, publishing, or safe rollout. |
Unknown | Not enough information to judge; gather specific facts rather than dumping a full checklist. |
Level 1 — Domain Status
Show a compact domain table.
| Domain | Status | Why it matters |
|---|
| Build / binary | pass / warn / block / unknown | AAB, signing, versionCode, target SDK, package name |
| Play Console packet | pass / warn / block / unknown | App access, app content, declarations, review notes |
| Data safety / permissions | pass / warn / block / unknown | Privacy policy, SDKs, sensitive permissions, data deletion |
| AI / UGC safety | pass / warn / block / unknown | AI reporting, prohibited outputs, moderation, minors |
| Store listing | pass / warn / block / unknown | Metadata policy, screenshots, icon, claims, localization |
| Testing / release tracks | pass / warn / block / unknown | Internal/closed/open/production, 12 testers if applicable |
| Billing / subscriptions | pass / warn / block / unknown | Play Billing, entitlements, pricing, cancellation |
| Accessibility / age rating | pass / warn / block / unknown | Target audience, rating, Families, Android accessibility |
| Sensitive categories | pass / warn / block / unknown | Health, finance, government, VPN, prediction markets, etc. |
| Quality / rollout | pass / warn / block / unknown | Pre-launch, crashes, ANRs, staged rollout plan |
Level 2 — Open only relevant domain skills
Use the specialized skill only when that domain is relevant or blocked:
android-expo-eas-build-binary-readinessandroid-play-submission-packet-readinessandroid-play-data-safety-permissions-readinessandroid-play-ai-ugc-safety-readinessandroid-play-store-listing-metadata-readinessandroid-play-testing-track-readinessandroid-play-iap-subscriptions-readinessandroid-play-accessibility-age-rating-readinessandroid-play-sensitive-category-readinessandroid-play-quality-prelaunch-readinessandroid-play-console-eas-automation-bridge
Level 3 — Remediation plan
For every blocker, provide:
| Field | Required content |
|---|
| Blocker | What will stop submission/review/release |
| Evidence | File, setting, policy form, screenshot, build log, or missing information |
| Fix | Concrete change |
| Verification | How to prove it is fixed |
| Automation | EAS / Play Console / API command only if appropriate |
| Owner | Engineering / product / legal / design / support |
Inputs to Gather
Ask only for the missing facts needed to produce a verdict. When possible, infer from project files.
Important facts:
- Is this a new app or an update?
- Is the Play developer account personal or organization?
- Is the app already granted production access?
- Does the app use Expo managed/prebuild/bare workflow?
- What Expo SDK and React Native version are used?
- Is the release built with EAS Build or local Gradle?
- Does the app use EAS Update / OTA updates?
- Does the app use AI-generated content?
- Does the app include UGC, chat, social, comments, profiles, or public sharing?
- Does the app require login, subscription, QR code, invite, external hardware, or seeded data?
- Does it collect personal/sensitive data?
- Does it request contacts, location, camera, microphone, photos/videos, health, SMS/call logs, accessibility, VPN, background location, exact alarm, notification, or foreground service permissions?
- Does it sell digital goods/subscriptions?
- Is it health, finance, government, VPN, news, dating/matchmaking, gambling/contest, prediction market, children/family, crypto, or regulated?
Hard Blockers
Treat these as likely blockers until proven otherwise:
- No unique
android.package / application ID.
- Duplicate or non-incremented
android.versionCode for a new upload.
- Wrong artifact for Play production: APK where AAB is expected for ordinary Play distribution.
- Target SDK below current Google Play requirement for new apps/updates.
- App not signed with valid upload key / Play App Signing not configured.
- Reviewers cannot access locked features, subscription content, seeded AI features, or account-only flows.
- Data Safety section missing, inaccurate, or inconsistent with app behavior or privacy policy.
- Privacy policy missing, PDF-only, geofenced, inactive, not clearly labeled, or not in-app when required.
- App collects sensitive data without prominent disclosure/consent where required.
- Broad contacts/location/photo/video/file permissions when system picker or narrower API is sufficient.
- Generative AI app lacks in-app offensive-content reporting or flagging.
- UGC app lacks terms, moderation, reporting, blocking, or child-safety safeguards.
- Digital goods/subscriptions sold outside Google Play Billing without a valid regional/program exception.
- Content rating or target audience is inaccurate.
- New personal developer account has not completed required closed testing before production access.
- App crashes, hangs, or cannot complete its core user journey.
- Metadata makes misleading AI, health, finance, privacy, security, or earning claims.
High-Risk Warnings
Treat these as warnings that may require specialized review:
- AI companion, AI therapy, AI medical advice, AI legal/financial advice, AI image/video generation, deepfake-like features, or real-person voice/video synthesis.
- Children or teens can access AI/UGC/social features.
- App has anonymous/random chat or matching features.
- App uses health data, Health Connect, wearable data, mental health content, or symptom analysis.
- App uses Accessibility API for automation, screen reading, interaction with other apps, or agentic workflows.
- App uses foreground services, background location, exact alarm, SMS/call logs, contacts, or broad storage access.
- App offers subscriptions, free trials, token packs, credits, or usage-based AI consumables.
- App uses Firebase/Google Sign-In/OAuth and certificate fingerprints may differ between debug, upload, and Play app-signing keys.
- App relies on external backend services that may not be live during review.
Current Google Play Policy Watch Items
Always verify current docs, but treat these as important current themes:
- New apps and app updates currently need a recent target API level; check the official target API level page before submission.
- New personal developer accounts may need a closed test with at least 12 opted-in testers for 14 continuous days before applying for production access.
- Google Play AI-generated content policy requires in-app reporting/flagging for apps that generate AI content.
- Developers are responsible for AI outputs that violate restricted content, deceptive behavior, child safety, or harassment policies.
- Data Safety declarations must include data handled by third-party SDKs/libraries.
- All apps need a privacy policy, even when no personal/sensitive data is collected.
- April 2026 updates introduced a Contacts Permissions policy and recommended narrower contacts/location access patterns.
- October 2025 updates added or clarified requirements around age-restricted features, health, accessibility API, subscriptions, malware terminology, and financial features declarations.
Output Format
Use this format by default.
Verdict
Not ready / Probably ready / Ready to submit / Unknown
One sentence explaining the verdict.
Domain status
| Domain | Status | Notes |
|---|
| Build / binary | block | versionCode unknown and target SDK not verified |
| Play Console packet | warn | Review access likely needs seeded demo account |
| AI / UGC safety | block | AI app has no in-app content reporting |
Blockers
| Blocker | Fix | Verification | Owner |
|---|
| AI-generated content lacks in-app report flow | Add report/flag action in AI output UI and moderation queue | Reviewer can report an output without leaving app | Product + Engineering |
Next actions
- Do the smallest set of actions that removes blockers.
- Then run the relevant EAS/Play Console automation bridge.
- Do not submit to production until blockers are cleared.
Rules for Automation
- Never mutate Play Console or submit a production release without explicit user approval.
- Prefer read-only inspection first.
- Prefer internal testing before closed/open/production if the account and app state are unknown.
- Use staged rollout for meaningful production updates unless the change is tiny and low-risk.
- Do not use automation to bypass unresolved policy issues.
Final Gate
Before recommending production rollout, confirm:
- AAB exists and is signed correctly.
- Target SDK is compliant.
versionCode is unique and incremented.- Play App Signing / upload key / API fingerprints are understood.
- Store listing is complete.
- App content declarations are complete.
- Data Safety and privacy policy match actual behavior.
- App access instructions are reusable and valid.
- AI/UGC/reporting/moderation requirements are satisfied if applicable.
- Billing and subscriptions are correct if applicable.
- Content rating and target audience are accurate.
- Pre-launch report and critical QA are clean enough.
- Rollout plan is staged or otherwise justified.
