-
-
Notifications
You must be signed in to change notification settings - Fork 0
Features
Nick edited this page Jul 28, 2026
·
1 revision
- CodeMirror 6 Markdown editor with a serif typeface optimized for long reading
- Manual save (
Ctrl+Sor Save button); unsaved-change indicator with confirm-before-close - Session undo and redo
- Formatting toolbar, font selector, find and replace
- Light and dark themes (app-wide, persisted)
- Selection highlight persists when focus moves to chat or Smart Advisor panels
- Reader Mode — clean two-page spread view of the full manuscript
- Global Search + Replace (
Ctrl+Shift+F) across all project files - Thesaurus (right-click a word, powered by Datamuse API)
- Spelling corrections in the right-click menu (offline Hunspell dictionary)
- UI font scale (Default / Larger / Larger+ / Largest)
- Text-entry sizing (separate, more aggressive scaling for chat/profile inputs)
Character, Relationship, Location, Lore. Each has its own section template.
Profile sections are made of trait blocks with importance levels:
- trait: observant, punctual, eloquent
description: "She is the textbook example of someone always on time."
importance: core| Level | Meaning |
|---|---|
| Core | Central to identity. Always sent to AI |
| Present | Regularly relevant. Sent when character is in scene |
| Background | Exists in canon but rarely surfaced |
| Contextual | Situational. Sent only when explicitly attached |
| Hidden | Writer-only. Never sent to AI; AI expresses as subtext only |
- Section summary generation
- Full profile summary (synthesizes all sections, weighted by importance)
- "How AI uses this" preview (sparkles popover per trait)
- AI Trim (rewrites wordy traits to tighter targets)
- AI Importance Audit (flags importance-level mismatches)
- Profile Builder chat (Chat, Refine, Extract Traits, Check Consistency modes)
- Import and fork profiles from other projects
Three pass categories with inline highlights:
- Readability — sentence complexity, passive voice, adverb density
- Structure — pacing, scene breaks, chapter flow
- Context — character consistency, profile adherence (requires attached profiles)
Features:
- Selection-scoped passes (highlight a paragraph, run on selection only)
- Eight creative-transformation modifiers (Default / Rewrite / Expand / Shorten / Describe / Rephrase / Add Sensory Detail / Change Tone)
- Word-level diff display showing exactly what each suggestion changes
- Issue ceiling calibrated by draft stage (polished: 0-5, revision: 10-25, raw: 30-50)
- In-context suggestions (improves the passage in its actual context, not isolation)
- Multi-turn chat panel for conversational AI help
- Context chip attachment (Summary / Traits / Overview / Details per chip)
- Chapter and scene summaries as context chips
- Markdown-preserving copy from AI chat bubbles
- Continuation mode (first word is the next word of the story)
See The Pipeline for the full autonomous production system.
Key capabilities:
- 11-phase resumable pipeline with human approval gates
- Post-critics revision loop (max 2 retries per chapter)
- User input overrides for any phase
- Auto-run mode with configurable delay
- Per-phase model routing
- Pipeline Companion chat with Apply to Brief + Re-run Phase
- Output Library with browsable artifact catalog
- Full manuscript export
- Dated snapshot export
- Formats: Markdown, TXT, DOCX, EPUB
- Optional inclusion of summaries, notes, and profiles
- Multi-book projects with shared canonical profiles
- Per-book character arc files
- Series-canonical profiles in
profiles/
- Project-completion gauge against targets from outline YAML frontmatter
- Per-segment breakdown: Manuscript 50% / Outline 10% / Profiles 30% / Notes 10%
- Daily goal tracker with 7-day sparkline
- Configurable skill levels (Newbie 500 to Professional 10,000 words/day)
- Day rollover (Midnight or Night Owl 4 AM)
- Dedicated planning screen with collapsible section cards
- YAML frontmatter for targets (word count, expected characters/locations/lore)
- Auto-syncs to
notes/outline.mdfor pipeline integration - Raw view escape to Markdown editor
- 22 LLM providers with Recommended Models panel
- Per-phase model routing (Primary / Critic / Planner)
- Tier filter (free / budget / standard / premium)
- Text-only model filter
- Starred/favorite models
- Content mode (General / Mature / Explicit)
- Writing skill level and daily goals
- Theme (light / dark)
- Interface size scaling
Welcome to the Open-Write Studio wiki!
This wiki covers everything you need to know about Open-Write — from installation to architecture to contributing.
New here? Start with Getting Started.