1. Model Processing and Training
Draftsteady does not use your private writing to train models. Generation sends prompts and selected manuscript context through OpenRouter to the chosen model provider. Provider retention and training practices depend on the provider and its current policies; Draftsteady does not guarantee zero retention or no training by every upstream provider.
2. Data We Collect & Why
We collect: (a) Account identity: Name, email, and Google ID (if OAuth is used) for authentication; (b) User content: Novels, chapter manuscripts, lorebook entries, and snapshots explicitly saved by you; (c) Usage metrics: Token consumption and model selection solely for service operation and reliability controls.
4. Third-Party Subprocessors
OpenRouter and the selected model provider process generation inputs and outputs. Google Identity Services verifies sign-in. The hosting platform stores account data and manuscripts and processes operational diagnostics. Each provider has its own applicable terms and privacy policies.
5. Your Rights: GDPR (EU) & CCPA (California)
You can export saved account data and delete your account through the application. Browser-only chat and recovery drafts are separate from server exports and deletion; manage these in your browser. Backups may retain previously saved records until their configured retention period ends. Contact the operator for access, correction, or deletion requests under applicable law.
6. Privacy Officer Contact
For privacy inquiries, use the contact form prototype to view the existing contact details. Delivery through that form is simulated. The operator must confirm a working privacy contact before public launch; no new domain mailbox is represented as operational.