Privacy

Last updated: 24 April 2026

This page describes what BookForge collects, why, and what it does not. It's written to be read — not to be read past.

Short version: Your manuscripts stay on your machine. Your AI calls go through your own Claude Code subscription, billed directly to your Anthropic account. What we store on our server is limited to what we need to run your account and understand whether the beta is working.

1. Who we are

BookForge is operated by Synaptrix AI ("we", "us"). Contact: hello@synaptrixai.com.

2. What stays on your machine

Your manuscripts and project files. Every .docx, .pdf, image, and working document lives in a folder on your disk. We never upload them.
Your Claude conversations. BookForge spawns Anthropic's Claude Code CLI locally with your credentials. The API calls go from your machine to Anthropic directly, not via our server.
Your local database. Pipeline state, chapter metadata, and artifact registry sit in a SQLite file on your disk.

3. What we do store on our server

To run your account and the beta, we store the following in our server database:

Account data: email address, password hash (argon2id — not the plaintext), display name (optional), sign-up date, account tier and role.
Pipeline telemetry: for each step your BookForge run executes, we record: step ID, skill name, status (running / completed / failed / etc.), duration, input/output token counts, cost in USD, model used, error message (if failed), and an anonymous per-user hash of the project and chapter identifiers. We do not receive manuscript text, file names, or file paths.
Crash reports: if something crashes unexpectedly, the app sends a scrubbed stack trace so we can fix the bug. Filesystem paths are replaced with <HOME> before transmission.
Feedback you submit: anything you send through the "Feedback" button in the app.
Server logs: standard web server logs (IP address, request path, timestamp, status code). Retained up to 30 days.

4. Why we collect it

Account data is how we know it's you when you sign in.
Telemetry tells us which pipeline steps are working, which are slow, and which are failing — so we can fix the hardest problems first during the beta.
Crash reports let us see bugs we'd otherwise never hear about.
Feedback is the most direct signal of what you need from us.

5. Who has access

You, through your account.
A small number of Synaptrix AI staff, for support and debugging.
No third-party analytics, no ad networks, no "trackers".

6. Where data lives

Our server is hosted in the EU. User data is stored in Postgres on the same server, with daily on-disk backups kept for 7 days.

7. Your rights

You can:

Request a copy of what we have about you.
Ask us to correct anything that's wrong.
Ask us to delete your account and data (within 30 days).

Email hello@synaptrixai.com to exercise any of these rights. We will confirm within 30 days.

8. Cookies

This website does not set cookies. The desktop app stores your login session token in an OS-encrypted file in its local data directory.

9. Anthropic's privacy

When BookForge runs a pipeline step, Claude Code on your machine sends the prompt to Anthropic. Anthropic's data handling is governed by Anthropic's privacy policy. We are not a party to that data flow.

10. Changes

If we make a material change to this policy, we'll send a notice to registered users before it takes effect.