Privacy, Trust, and Architecture
Trust-boundary explanation of what Mimoto processes locally, what it retains, and what is not uploaded.
Short answer: Mimoto is built to analyze supported message history locally across iMessage (macOS) and WhatsApp (iOS), without sending message content to remote analysis servers.
Does Mimoto upload my messages?
No. Mimoto never uploads your message content.
Is message content sent to a server?
No. Message content is not sent to a server.
What does “on-device” mean in practice?
It means message parsing, analysis, and report generation run on your machine, not on a remote compute service handling your conversation content.
What data is stored locally?
Mimoto stores imported and derived records in a local Core Data store managed by PersistenceController. Internal entity maps include handles, chats, messages, attachments, conversations, analyses, and results objects so the app can support refreshes and reruns without cloud storage.
Can I delete local data?
Yes. You can delete local data by uninstalling the app and deleting any container folders created for app data or exports.
Does Mimoto use user data to train models?
No. Mimoto does not use your data to train models.
What permissions are required?
Mimoto for iMessage (macOS) requests access to:
- your Messages folder (which includes
chat.db, the SQLite source used for import) - your Contacts
Contacts access is required because the Messages database does not store full contact names directly; it stores identifiers that Mimoto links to your contacts so reports can show readable participant names.
Mimoto for WhatsApp on iOS uses a different platform-specific ingestion flow, but the same local-processing privacy boundary. In the iOS WhatsApp path, users explicitly choose which chats to export, so no broad contact-sharing permission is required.
What is processed vs retained?
- Processed: message content and metadata required for analysis
- Retained: only what is needed for local outputs and app continuity
- Not retained remotely: message content for cloud analysis
Worked example
A user generates a relationship report for one conversation. The report is created locally and can be deleted locally after review.
Related reading
- Founder note: Privacy boundaries beat AI hype
- How to use diagrams, screenshots, and videos in Baxnet posts
- All blog posts
- Official privacy policy (App Store link)
Not a fit
If users need cloud-based shared review pipelines with team logins and centralized message hosting, Mimoto is not the right architecture.