Automating Google Play Store Listings with Claude Code

Maintaining Google Play Store listings across 22+ apps is the kind of work that compounds invisibly. Each app has a title, short description, full description, screenshots, content rating, Data Safety declaration, and release notes per release. Multiply by 22+ apps and 2+ releases per month and you're spending hours per week on copy work.

Here's how I removed that overhead using PlayCraft, a Claude Code plugin built specifically for Play Store publishing.

What the manual process actually costs

Per app per release, the Play Store work is:

• Update short description if features changed
• Update full description bullet points
• Generate release notes from commits
• Re-verify Data Safety declaration
• Re-verify content rating
• Update screenshots if UI changed
• Submit and wait for review

Most of these are 5–10 minute tasks. For 22 apps releasing monthly, that's 3–4 hours of repetitive copy work — work that has no engineering leverage and that I was getting worse at over time because I was doing it under fatigue.

What PlayCraft does

PlayCraft is a Claude Code plugin that encodes the full Play Store workflow as a structured set of skills. Each app has a metadata file describing its features, target audience, and key value propositions. PlayCraft reads this, plus the recent commit history, and produces:

• A polished store description matching the app's positioning
• Release notes in the right format (under 500 characters, user-facing language)
• A Data Safety declaration based on what permissions the app uses
• A content rating questionnaire response based on app type
• A localized version of all the above for each target market

The output is structured — JSON or markdown that maps directly to Play Console fields, not narrative text that I have to copy-paste and reformat.

The skill structure

Each Play Store task is a discrete skill:

• code Copy

  generate-release-notes

  — diffs the last release tag against HEAD, summarizes user-facing changes
• code Copy

  update-store-description

  — reads metadata + current Play Store state, suggests updates
• code Copy

  data-safety-declaration

  — analyzes manifest permissions, generates declaration
• code Copy

  content-rating-questionnaire

  — produces ESRB/IARC-compatible responses
• code Copy

  localize-listing

  — translates listing to target languages with marketing-appropriate tone

Skills compose. A full release flow chains all of them through a single Claude Code invocation.

Where it earns its place

The single highest-value automation is release notes generation. The manual process — scroll through commits, identify user-facing changes, write 3 bullet points in product language — takes 10–15 minutes per app per release. PlayCraft does it in 30 seconds and the output is consistently better than what I'd write tired at 11pm.

Second highest: Data Safety declaration. This requires going through every permission your app requests and declaring what data is collected, why, and whether it's shared. Getting this wrong is a Play Store rejection. Doing it manually for 22 apps is error-prone. Doing it automatically from the manifest + a structured questionnaire is reliable.

Where it doesn't replace human judgment

Store listing copy still needs human review. AI-generated copy is competent but generic. The positioning, the unique value prop, the emotional pull of the description — that's still a human-writing task. PlayCraft handles the structure and consistency; I edit the voice.

Screenshots are not in scope. Visual design still requires human work.

The actual savings

Across 22 apps releasing monthly:

• Release notes: 22 × 15 min = 5.5 hours saved per cycle
• Data Safety re-verification: 22 × 5 min = 2 hours saved
• Store description updates (when needed): variable, but ~1 hour saved
• Total: ~8 hours per month of repetitive Play Store work eliminated

That's a full work day per month, every month, recovered.

Why this is plugin-shaped, not script-shaped

A script could generate release notes from git log. The plugin layer matters because:

• It handles the full workflow, not one step
• It composes with other Claude Code skills (changelog generation, version bumping)
• It runs inside the same Claude Code session where I'm already working
• It's installable across all 22 apps via the plugin registry — no per-repo setup

Encoding Play Store workflow as a Claude Code plugin made the work scalable. Encoding it as a script per app would have produced 22 scripts that all drift apart.

Pattern for other tooling

The same pattern applies to any repetitive product work:

1. Identify the manual workflow with the highest time cost and lowest engineering leverage
2. Map each step to a discrete skill with clear inputs and outputs
3. Encode the skills as a Claude Code plugin with composable chains
4. Install once, invoke from any project

For solo developers across a portfolio of apps, this kind of tooling is the difference between "managing 5 apps comfortably" and "managing 22 apps without burning out."

PlayCraft is one example. The same approach applies to privacy policies (PrivacyPilot), Android build workflows (DroidForge), and Kotlin language tooling (KotlinSense). Each one removes a specific category of repetitive work permanently.

The compound savings across all of them is what makes solo fleet management feasible.