The Keyboard free beta downloads the prompt catalog over HTTPS.
Recent and favorite identifiers are stored only on the device. A selected prompt is inserted into the destination field only when you choose it.
Insertion is blocked in password fields, and the IME does not read or transmit text that you type in the destination app.
Library beta
Library beta processes your email address for email magic link authentication and account management. When a sender enters a recipient email address for direct sharing, we also store and use that address to deliver the share to the Library Inbox associated with it.
We store your profile public name, private, directly shared, or public prompt titles and bodies, optional source URL and cover image, sync metadata such as revision and update time, and direct-share data including the recipient email address and delivery state.
Direct sharing is separate from public publishing. While a share is pending, the recipient signed in with the addressed email can view it in the Library Inbox. Saving it creates a private, independent copy in the recipient's account from the content at that time. Later sender edits, stopping the share, or deleting the source do not automatically change or delete that saved copy.
Published content can be viewed and saved by other users. An independent copy saved by a recipient remains after the source public prompt is deleted.
For reporting and moderation of public content, we process the report, the affected item, and its review and action status.
Chrome paid features
We collect email addresses and payment-related identifiers for purchase flow and entitlement sync.
Payments are processed with Stripe and required information is shared only as needed.
We record anonymous usage events needed for product improvement, such as Daily Prompt views, panel opens, tab views, locked feature clicks, prompt insertion success or failure, pricing page views, plan selection, purchase button clicks, and purchase flow return, cancellation, or error events.
For anonymous events, a browser-level random ID is hashed server-side before storage, and records are limited to fields such as session ID, locale, plan, free or paid access tier, surface, tab, catalog type, item ID, reason category, hour, weekday, and duration bucket.
We do not store prompt text, user input, generated content, page content, email addresses, raw anonymous IDs, or Stripe customer IDs in anonymous events.
When a paid user explicitly chooses backup, saved prompts and Collections are stored in BananaNL app data inside that user’s Google Drive. BananaNL servers do not store backup prompt bodies.
Data obtained from Google APIs is used only to provide and improve user-facing Google Drive Backup and Restore features, and is not used for advertising or unnecessary third-party disclosure.
BananaNL's use and transfer of information received from Google APIs adheres to the Chrome Web Store User Data Policy, including the Limited Use requirements.
Collected data is used for service delivery, purchase confirmation, support, and product improvement.
We do not disclose data to third parties beyond what is required by law or service operation.
