Skip to content
WebViewGold Vibe Coding to App Store
English
Get WebViewGold

Claude Code to the app stores

Setting up Claude Code for a team: shared memory, permissions, review rules and secrets

One developer with Claude Code needs good habits. A team needs agreements, because five people with five private setups produce five kinds of code, and nobody knows what the agent was allowed to do. Four conventions cover most of it, and all of them fit into the repository.

Free 15-minute call with our team, or call +49 761 519 751 50 (24/7, English + German)

  • Approval guarantee – or money back within 14 days
  • No coding needed
  • One-time payment, free updates
  • 24/7 human support
  • Made in Germany since 2014 · ISO 27001

Four agreements

Shared project memory

One CLAUDE.md in the repository that every session reads, maintained like code.

Permission settings

A common baseline for what the agent may do without asking, what it must ask for and what is blocked.

Review rules

A written list of changes that never ship without a person reading them.

Secrets handling

Clear answers to where keys live and which of them the agent can ever see.

Shared project memory

The project memory file tells the agent how the app is built, tested and released, and for a store app also which rules apply inside the native shell. Keep one version in the repository and change it through reviewed commits. Personal preferences, such as how much the agent should explain, belong in each person's own settings, not in the shared file. The CLAUDE.md template is a starting point for the store-related part.

Two habits keep the file alive. Add a line whenever the agent makes the same mistake twice or a store rejection teaches you something. And delete lines that no longer apply, because a long file full of old rules is followed less well than a short current one.

A permission baseline

ActionSettingReason
Read code, run tests, linter and buildAllowedSafe and needed constantly; prompts here only train people to click yes
Edit files on a working branchAllowedEvery change is reviewed before it is merged
Install or upgrade dependenciesAskA new package is a security and a size decision
Push, deploy, run migrationsAsk, or reserved for peopleDeploys reach the installed apps immediately
Read environment files, keys and signing materialBlockedThe agent does not need them for its work
Commands against production systemsBlockedMistakes there cannot be undone with git

What a person always reviews

  • Login, sessions and password reset.
  • Payments, subscriptions and anything that decides what a user has paid for.
  • Account deletion and data export, because they destroy or disclose data.
  • Database access rules and migrations.
  • The privacy policy text and every answer that goes into a store form.
  • Native configuration: permissions, app identifiers, the URL the app loads.
  • New dependencies and new third-party scripts.
  • Changes to the memory file and to the permission settings themselves.

Secrets, and where the rules live

Commit the permission baseline with the project, so it applies to everyone and to automated runs, and agree that individuals may make it stricter for themselves, never looser. For secrets the rule is short: none in the repository, none in the memory file, none in prompts. The agent works against a development environment with test keys. Production keys live in the settings of the hosting provider, signing keys and store logins with named people. If a key shows up in a session or a commit anyway, rotate it first and tidy up afterwards; the secrets guide has the review routine.

Our team sets up these conventions in workshops on real codebases, with playbooks for CLAUDE.md, MCP and code review and workflows aligned with ISO 27001. The scope depends on your team, so book a free call to talk it through.

Frequently asked questions

Should every developer have a personal CLAUDE.md?

Personal preferences can live in personal settings. Everything that affects how the product is built belongs in the shared file, so all sessions follow the same rules.

May the agent push to the main branch?

For a published app, no. Every deploy reaches users at once, so changes should arrive through a reviewed pull request.

Who is responsible for code the agent wrote?

The person who merges it. Agree on that explicitly, because it is what makes people read the diff.

How do we onboard a new team member?

Give them the repository and one small task to complete with the agent under the shared rules. The memory file and the permission baseline carry most of what used to be passed on by word of mouth.

Three ways to get your app into the stores

Same result, different amount of work on your side. All three come with the approval guarantee and 24/7 human support.

Do it yourself

$109 per platform, one-time

The WebViewGold license: a finished Xcode or Android Studio project. Paste your URL, set name and icon, switch on features, build and submit. Free updates forever.

See plans on the main page

Cloud Builder

$349 per license, one-time add-on

No Mac, no Xcode, no Android Studio: configure in the browser, we build both apps in the cloud and hand them to App Store Connect and the Play Console.

How the Cloud Builder works Try the free demo

Full Service

$698 per platform $798

Send us your URL and icon. Our team configures, builds and delivers your app within 72 hours, white-label, with source projects. Both platforms for $1,158.

Full Service details Powered by WeSetupYourWebViewApp

Winter Sale prices valid until December 31, 2026.

Talk it through with our team

Free 15-minute call: your app, your tools, your next step

Rather start right away? See plans and pricing or hand it over with Full Service.