Local-first manuscript processing
Your manuscript stays local for authoring and Word generation
Manuscript content stays on your device for authoring, review and DOCX generation. Account sign-in, entitlement checks and future optional cloud services may communicate with Barker Graftworks systems, but manuscript text is not required for the core book-building workflow.
What stays local in Stage 1
Project text, images, Markdown exports and generated Word files are stored or produced on the customer’s device. They are not uploaded merely to make a DOCX or check an entitlement.
What the account service does know
The backend stores the minimum non-manuscript information needed for accounts and commercial access: identity, opaque book identifiers, sessions, payment references and entitlement state.
No manuscript text in analytics
Analytics must never receive manuscript text, book titles, subtitles, chapter titles, section content, image content or ISBNs. Measurement is limited to product and commercial events such as a page view or completed signup.
AI optional. Never AI imposed.
The core manuscript-processing workflow does not use generative AI to write, rewrite, edit or train on customer manuscripts. Customers remain free to export Markdown and use an external AI tool on their own terms.