ScriptaDraft is an AI-powered manuscript development platform designed for academic researchers. It helps you structure, refine, and stress-test manuscripts — with AI-assisted writing improvement, simulated peer review, scheme-aware scoring, and reference discovery. It is a writing and review aid, not a content generator and not a submission system.

ScriptaDraft is built for academic and clinical researchers preparing journal manuscripts. It is particularly suited to those targeting high-impact journals such as Nature, Science, Cell, and the Blood family, though it can support any journal and article type.

Built-in schemes cover Nature, Nature Medicine, Nature Communications, Science, Cell, Immunity, Blood and Blood Advances, each with its article types' word and figure limits and review criteria. Always confirm limits against the journal's current author guidelines before submission. For any other journal, choose “Custom / Other journal” and paste in its guidelines.

ScriptaDraft's AI reviewer simulates the perspective of an expert peer reviewer for your target journal and article type. It assesses your section against that journal's review criteria, identifies weaknesses, and provides structured feedback — similar to what you might receive in peer review, but available instantly during drafting.

The peer review simulation generates a structured review of your manuscript from a simulated reviewer's perspective — per-criterion strengths and weaknesses, narrative feedback, and an editorial recommendation (Accept, Minor revision, Major revision or Reject). You can choose from different reviewer lenses (e.g. methodologist, clinical, sceptical).

The AI assessments are calibrated to each journal's review criteria and editorial bar, but they are simulations — not guarantees of reviewer opinion. Feedback should be treated as one informed perspective to help you strengthen your manuscript, not as a prediction of the editorial decision. Always seek input from experienced colleagues alongside AI feedback.

Journal and publisher policies on AI use are evolving and vary. Following COPE and ICMJE guidance, major publishers do not accept AI tools as authors and expect authors to disclose how AI was used; some also restrict AI-generated images. ScriptaDraft is designed as a writing and review aid — it helps you improve your own draft and stress-test it before submission, not generate substantive scientific content. Compliance with your target journal's and your institution's policies is your responsibility. See our Acceptable Use Policy for what ScriptaDraft must not be used for.

ScriptaDraft routes AI requests to frontier models on ZDR-compliant infrastructure. Beta and Free accounts use a server-side key — no personal API key is required, and the right model is selected automatically for each task (heavier reasoning for assessments and simulated reviews; faster models for chat and lighter tasks).

Live EndNote citations carry hidden citation data inside your Microsoft Word document. When you paste from Word in Chrome or Edge, ScriptaDraft keeps that data, and the citations are still live when you copy the text back into Word.

Safari does not pass embedded citation data to web pages, so citations pasted in Safari arrive as plain text (for example, superscript numbers). Your text and formatting are unaffected.

To keep your citations live, either paste from Word using Chrome or Edge, or, in Word, choose EndNote → Convert Citations and Bibliography → Convert to Unformatted Citations before copying. Citations then paste as text such as {Smith, 2020 #12}; after copying the text back into Word, choose Update Citations and Bibliography to restore them.

You don’t need to copy your reference list back into Word: once your citations are back in Word, Update Citations and Bibliography rebuilds it for you.

Yes. Your manuscript content is stored in an AES-256-GCM encrypted database on secured infrastructure. It is not shared with other users, not used for training AI models, and not accessible to third parties.

All managed accounts (Beta and Free) use Zero Data Retention (ZDR). Your manuscript content is processed transiently — it is not stored by the AI provider after the response is returned and is not used to train AI models.

No. ScriptaDraft is an independent product developed outside of any institutional or employment context. Your use as an individual does not create any relationship between the developer and your employer or institution, and does not constitute organisational adoption, endorsement, or licence.

AI processing is measured in tokens — roughly, the amount of text sent to and received from the AI. Beta accounts include a 1,000,000 token allowance, sufficient for 10+ full manuscript reviews depending on manuscript size and the level of AI assistance you utilise across other features. Demo accounts include 100,000 tokens. You can track your usage at any time in Settings → tokens.

Not during the current beta. All accounts use the managed, ZDR-compliant infrastructure described above, and the right frontier model is selected automatically for each task. We may offer a BYOK (bring your own key) option in future once the managed-tier configuration has stabilised.

ScriptaDraft works in all modern browsers — Chrome, Safari, Firefox, and Edge. Chrome or Safari are recommended for the best experience. Mobile browsers are not currently optimised.

Beta accounts have a 1,000,000 token allowance and full access to all features. Free accounts have a 100,000 token allowance and are intended for short-term evaluation.

Use the bug report button (available throughout the app) to send feedback directly to the development team. You can also email [email protected]. All feedback is reviewed and genuinely appreciated.

We intend to migrate beta user data to the production platform. However, as this is beta software, we cannot guarantee data persistence — please keep your own copies of important manuscript documents.