Frequently Asked.
Honest answers about credits, ownership, AI training, the imprint, and how the team works.
Credits are how the AI tools are metered. Different tools cost different amounts. Bea (copy editor) is 5 credits per chapter, the Professor is 2 per query, Find an Agent is 2, Jules (rewrite partner) is 2 to plan, 6 to rewrite, 4 to iterate, the Gauntlet runs 5 to 60 depending on scope. Every tool shows its cost before you commit.
New accounts come with a starting balance. You can top up via the Gift Center. Credits don't expire.
The tool will stop before charging you and tell you exactly how many credits the operation needs. Your work isn't lost. Top up, run again.
Yes, on unused credits within 30 days of purchase. Email drterryoroszi@emerging-tech-lab.com with your account email and we'll process it.
No. Your manuscript is sent to Anthropic's API for processing only. Anthropic's API terms forbid training on customer inputs. Your text is not used to train any model. We do not retain your manuscript on our servers beyond what's needed to run the tool you asked for.
You do. Every word the tools return is yours to use, edit, sell, publish, or burn. We claim no rights to your output.
What we ask in return is honest attribution. If a tool helped you ship a manuscript, you don't have to credit Greylander Press by name — but you do agree not to misrepresent the work as wholly AI-untouched, and you agree to follow the PAID framework if you publish through the GP imprint.
PAID is Dr. Terry Oroszi's framework for rigorous AI use, published in Forbes: Position, Audit, Interrogate, Demand. Standard AI models have a built-in bias toward sycophancy, they're trained to agree with you, validate your ideas, and hedge in comforting, overly polite language. That makes for a pleasant chat. It's useless for an author who needs a rigorous, objective read on a manuscript.
PAID counters that directly: position the AI's role honestly, audit what it actually produces against what's true, interrogate hedges and gaps, demand specifics over generalities. Greylander Press's tools enforce it architecturally, not by asking the AI nicely to be honest. The Gauntlet's evaluation engine detects hedged, soft, or compensatory framing and forces itself to regenerate until the output meets a fixed, uncompromising honesty standard. That means you should expect a critical, unvarnished assessment of your draft, not a flattering one. Unfiltered truth is the product, not your ego. Full definition on the About page.
No. The tools run on Anthropic's Claude Sonnet 4.6 with structured prompts built specifically for fiction craft. Each character (Grey, Bea, Margo, Jaque, the Professor) has a tuned system prompt with rules against sycophancy, hedging, and house-style violations. The tools are designed to disagree with you when you're wrong.
The tools are the AI characters: Grey, Bea, Margo, Jaque, the Professor, plus the Gauntlet and Honest Book Reviews. Anyone with credits can use them. What you write with the tools is yours, and you can publish wherever you want.
The imprint is Greylander Press as a publisher. We accept manuscripts (most through the Gauntlet diagnostic), produce books to a high editorial standard, and distribute through KDP and ACX. The imprint is selective. The tools are open.
No. A vanity press makes its money charging authors a flat fee to print anything, no editorial standard required, so it profits the same whether the book is good or not. That's not how this works. The tools run on metered credits, the same rate whether your manuscript is sharp or a mess, there's no incentive on our end either way.
The imprint itself is selective, not automatic. Submission review is free, we assess your manuscript before any fee discussion, and if it's not a fit we tell you clearly. Most manuscripts come in through the Gauntlet diagnostic, and we turn work down. What we do accept, we handle the ISBN assignment, cover development, KDP print and ebook formatting, and ACX audiobook placement for. You keep your rights either way, tools or imprint.
No. The tools are yours to use however you want — query traditional agents, self-publish on KDP, give your manuscript to a friend, never publish at all.
The Gauntlet is a private manuscript diagnostic for the author. Banded scoring, no public footprint. Run it on your own work to see where you stand.
Honest Book Reviews are public verdicts on books that have been submitted to GP and read by the multi-pass review engine. The reviewed catalog is published. Same anti-sycophancy standard, but the audience is readers, not the author.
They're the AI tools, given personas so it's clear what each one does:
Grey — your writing partner. Workshop tool: enrich, dialogue, continue, diagnose, rebuild.
Jules — rewrite partner. You tell her what's not working in plain English. She proposes a game plan, then rewrites only after you sign off.
Bea — copy editor. Grammar, punctuation, dialect-preserving line edits.
Margo — beta reader. Reactions, not edits. Where you got hooked, where you got bored.
Jaque — query coach. Polishes query letters, log lines, synopses, comp titles.
The Professor — research librarian. Fact-check, period accuracy, technical authenticity.
Each character has its own page, its own credit cost, and its own rules. They're not therapists. They will tell you when something is broken.
If you know what tool you want, use the tool. Prose Enricher, Structural Rebuild, and Add Dialogue each do one thing fast.
Use Jules when you can describe the problem but not the fix. She is the guided path: pick the chapter, check the boxes that match what bothers you, type a sentence in plain English. She reads the chapter, proposes a short game plan, and only rewrites after you sign off. You can iterate on her rewrite with "more of this, less of that," paragraph by paragraph.
Because "an AI tool that does six different things" is harder to use than "Grey, who writes alongside you" and "Bea, who copy edits." Personas tell you what role you're invoking. The character is a UI, not a claim.