Modeled on professional developmental editing. Powered by the AI model you choose.
This mirrors how a professional developmental editor works — understand the manuscript before judging it.
Builds reference documents — character lists, timelines, plot threads — before critiquing a single word.
Evaluates across 6 analytical dimensions tailored to your content type. Nothing is skipped.
Every issue is categorized by type and prioritized into one of three severity tiers.
Evaluation criteria shift based on your genre's conventions. What's a flaw in literary fiction may be a feature in thriller.
Each content type has its own issue categories, severity weighting, and evaluation criteria. 7 content types. 45+ genre configurations. 6 analytical dimensions per type.
Full-length narrative fiction with chapter structure
Compact narrative fiction under 20,000 words
Argumentative, analytical, or personal essays
Digital-first content with audience engagement focus
Feature film or TV script in standard format
Personal narrative based on real events
Research papers, theses, and scholarly articles
Not everything is critical. Our tiering prevents revision overwhelm.
Must fix before publication. Plot holes, contradictions, structural collapses — issues that undermine the work's foundation.
Materially weakens the work. Pacing problems, flat character arcs, inconsistent voice — things a developmental editor would flag.
Separates good from great. Prose rhythm, word choice, subtle thematic resonance — the work that elevates a manuscript.
You get a plan, not just a list of problems.
Address story-breaking issues first. Foundation before decoration.
Fix pacing, restructure scenes, reorder for narrative momentum.
Deepen character arcs, strengthen dialogue, unify narrative voice.
Line-level refinement. Sentence rhythm, word choice, clarity.
Final pass for consistency, fact-checking, and genre conventions.
Choose the kind of feedback you actually want.
Supportive and constructive. Celebrates strengths while guiding improvement.
Precise and analytical. Focuses on craft mechanics and structure.
Honest and collegial. The feedback you'd get in a good writing group.
Evaluates against genre conventions and reader expectations.
Direct and unsparing. What a no-nonsense editor would tell you.
Each preset comes with fine-tuning sliders for directness, detail level, and encouragement balance.
Galleys was developed alongside The Third Silence, a 300,000+ word fan completion of Patrick Rothfuss's Kingkiller Chronicle.
A good editor is never “satisfied” — and that's by design. Here's how to know when you're done.
Tier 1 issues deserve a decision, not obedience. They're flagged as story-breaking — contradictions, timeline breaks, plot holes — and because they're derived from the text itself, they will reappear on every re-analysis until the text changes. But you're still the author. If a “contradiction” is intentional, dismiss it. A conscious overrule is a legitimate final answer, and a dismissed issue stays dismissed on future runs.
Tier 2 and 3 notes are suggestions, not defects. These are judgment calls, and the editor deliberately raises only the ones it's confident about — which is why borderline observations come and go between runs. If a note doesn't resonate after you've genuinely considered it, skip it without guilt. The editor is a well-read second opinion, not an authority on your voice.
New issues on every re-run are normal. Fixing structural problems exposes prose issues that were hidden underneath; polishing prose can reveal small continuity gaps. The revision plan is iterative by design, not a one-shot checklist.
The stopping rule: you're done when Tier 1 is clear — every issue fixed or consciously dismissed — and the remaining notes are ones you've considered and disagree with. “Zero issues” is not the goal; no manuscript, including published bestsellers, would get a clean report.