Skip to content
v1.2.0 Latest

Magic System Workshop, scene-level contradiction checks, three multi-agent council surfaces, and a fully-local engine that ships smaller, faster, and tells you which handler answered each chat reply.

Writing & Editing

  • Theme detection across the manuscript scans for recurring motifs and themes and reports them with cleaner labels (no more "morning", "took", or "stood" cluttering the topic list).
  • Theme and motif drop detection uses real chapter-gap math instead of estimated shares, so quiet setup gaps no longer trigger false alarms.
  • Editorial report composer lays out motifs, themes, and voice drift findings in their own sections. The voice drift block only shows up when there is something to flag.
  • Lorekeeper popover in the editor: hover a flagged finding to see what triggered it without leaving the page.
  • Find and Replace built into the editor.
  • Breadcrumbs across the workspace so it is always clear where you are in deep nav.
  • Source badges on every consistency finding name which check produced it (scene contradictions, ProseGuard, Lore ML, Lorekeeper).
  • Unified strictness slider (off, low, medium, high) across every consistency panel. Replaces the older 4-step depth dropdown in Lorekeeper.

Worldbuilding

  • Magic System Workshop. Three new tools for building the mechanical content of your world: Ability Workshop (templates, name patterns, random generator), Stat Forge (archetype templates and balance heuristics), and Sheet Studio (template-driven character sheet editor with a bundled template library, replacing the old Sheet Designer).
  • Legendry duplicate merge: collapse two duplicate entries into one. Aliases, tags, content, and relationships all union into the surviving entry.
  • Entry type prediction: the engine reads the content you have started and suggests the right Legendry entry type (character, location, faction, etc.) when you create a new entry.

Story Structure

  • Scene-level contradiction checks catch knowledge violations (a character knowing something they should not), point-of-view violations (slipping into a non-POV character's head), timeline ordering issues, missing-presence and unannounced-presence problems, and dead character appearances.
  • Pronoun chains feed presence checks. If a character is referenced as "he" or "she" through a scene without being named, presence detection follows the chain and counts them as present.
  • Manuscript scan: theme detection runs across the whole book and reports real progress as it goes.

Councils and Studio

  • Three creative surfaces are now structured multi-agent debates. Each one runs a fixed roster of agents through several handler rounds, then a deterministic composer turns the accumulated work into clean drafts plus a full report. Nothing leaves the machine.
  • Worldbuilding Council. Archivist runs consistency and contradiction sweeps, Worldsmith runs name generation, Scholar pulls real-world research and synthesis. The composer produces 3 to 8 fully formed lore-entry drafts plus a council report.
  • Character Council. Psychologist runs entity extraction and consistency, Dramatist runs the insight scan, Relationship Analyst runs relationship and plot analysis. The composer produces 1 to 3 character drafts plus a character council report.
  • Creative Studio. Storyteller seeds the copy, Strategist pulls market and audience research, Analyst runs the marketing copy balance check at high strictness. The composer produces four marketing assets (book blurb, short blurb, tagline, elevator pitch) plus a Studio report.

Chat

  • Every chat message now routes through the intent classifier and dispatches to the right specialist handler. The bubble caption tells you which agent answered and which handler did the work, so it is always obvious where the reply came from.
  • Suggestion chips above the input let you pick the routing target before sending. Click a chip and the message goes straight to that handler, bypassing the classifier.
  • Confidence floor. When the classifier is not sure enough, the engine asks a clarifying question instead of guessing. Tunable in Settings.
  • Persona and stance pickers are gone. Voice now comes from the project-genre overlay, not from a per-message dropdown.
  • Smart bookmarks and saved-page analysis run on Divinity entity extraction now. Deterministic summaries and tag sets land instantly.

The Divinity Engine

  • Fully local. Every external LLM provider is gone. There are no API keys to set, no cloud calls, no per-message costs, and no internet dependency for the engine itself. Everything that used to dispatch to Anthropic, OpenRouter, or Ollama now runs through the bundled handlers on your machine.
  • Authored Library expanded by roughly 50 new bundled situation packs across the three council surfaces, marketing copy feedback, chat-dispatch routing replies, and a per-handler chat-summary pack for every routable Divinity handler. Every chat reply now names the routed handler and tells you which surface to run it from.
  • Voice-lint passes clean across all 220 bundled situations.
  • Genre overlays: 12 genres with multiple variants across 130+ situations. The voice the engine writes in matches your project's genre.
  • Intent classifier: chat suggestion chips guess what you are asking for and route to the right agent.
  • Visibility learning: rules you engage with stay surfaced. Rules you ignore drift down. Per-rule overrides if you want to force a rule on or off regardless of behavior.
  • Pack registry browser installs Authored Library expansion packs from inside Settings, Divinity Engine, Pack Manager.
  • Variant review queue: review and approve new Authored Library variants in one place.
  • Marketing copy balance check is a new Divinity handler that scores a piece of marketing prose for readability, density, hook strength, length fit, and pain-point coverage. Used automatically by the Creative Studio Analyst, and available in Etherforce for ad-hoc copy passes.
  • Smaller, faster Deep Analysis Pack. Coreference resolution swapped from a 1.93 GB Longformer model to a 362 MB distilled one. Roughly 16x faster on CPU, with about 1 F1 cost on OntoNotes for the size and speed gain.
  • Embedding and contradiction models swapped to multilingual e5 and mDeBERTa v3 for richer semantic similarity and better cross-language consistency catches.
  • Orchestration tunables surfaced in Settings, Performance: utility scorer thresholds and behavior-tree weights. Changes take effect without a restart.

Desktop Experience

  • Boot diagnostics. If the engine fails to start, the splash names the failing step and shows the underlying error instead of timing out silently.
  • Real error messages everywhere. No more silent fails. No more "something went wrong." The actual cause surfaces wherever you can act on it.
  • Background loads (panel mounts, doc switches, tab opens) surface inline if they hiccup, instead of popping the global error modal.
  • Tray menu no longer crashes when the main window closes before the tray.
  • Studio modal stays mounted when you open it from anywhere in the app. The section editor never strands you on a placeholder button.
  • Etherforce gets a per-handler benchmark runner. Pick a handler, choose the fixtures you want to run, capture a baseline, and watch live progress as each fixture runs against the new code path. Diffs and pass / improved / regressed status surface inline so you can spot drift before it ships.
v1.1.8

Legendry Studio workbench polish, smarter entity detection, and streaming resilience across Lagan and benchmarks.

Writing & Editing

  • Failed model responses now show a labeled Retry button. Retry restages attached images and resends your message without duplicating it in history.
  • Streams that cut off mid-response surface a clear error instead of spinning indefinitely.

Worldbuilding

  • Legendry Studio workbench: collapsible section rail (Ctrl \), collapsible assistant rail (Ctrl Shift \), and a collapsible metadata drawer (Ctrl E) that tucks Dossier Basics, Timeline, Classification, WorldSpell, and Naming inspectors out of the way. Studio opens fullscreen by default on first open and Ctrl S saves.
  • New always-visible meta strip on each Legendry entry: title, type/status pills, word count, drafted fraction, and category/author/timeline breadcrumb.
  • Entity Cleanup panel — purge polluted Legendry rows directly in the app instead of running a terminal script.
  • Entity confidence is scored by real signals (repetition, capitalization, grammatical role, suffix, common-word and heading-bleed penalties) instead of a flat 0.8. Candidates that match an existing Legendry title or alias are pinned at full confidence.
  • Stricter entity deduplication: worldbuilding names like "Garuda" and "Garuyda" are no longer merged into a single entity.
  • Deleting an entity fully removes it everywhere, so it can't resurrect on the next Legendry reload.
  • One unified entity sidebar shared between the writing view and the Legendry article view — confirming or denying in one place updates the other instantly.
  • New Settings slider for the entity re-scan threshold (default 200 characters) so you can tune how aggressively both views re-detect entities.
  • Smarter date confidence: strict era-labeled dates score near the top, loose one-offs are dropped automatically.

Model Integration

  • Per-project agent voice: Personality, Temperature, Max Tokens, and Search Results are now per-project sliders in Settings. Voice text is threaded into every chat agent (Ishvana, Hawken, Lagan, GameMaster, WorldKnowledge) as tone guidance.
  • Per-project Knowledge Filter to bound what Legendry entries each agent can pull from.
  • Lagan chat streams in real time with progressive tool calls, thinking deltas, heartbeats, and an abortable run. Perception summaries (web pages and long content) now stream as well.
  • YouTube transcription jobs are scoped to the active project and no longer bleed across worlds.
  • New Lorekeeper Settings tab exposing enrichment concurrency, anomaly threshold, save/activity timeouts, retry backoff, consultation depth, and relationship limits. Recent Activity loads automatically when the panel opens.
  • Empty-stream resilience: if Anthropic, Ollama, or OpenRouter close a stream before delivering content, Ishvana retries automatically instead of returning a blank message.
  • Ollama "not reachable" messaging now names the exact base URL that failed instead of surfacing a raw exception.

Benchmarks & Models

  • Benchmark grader stopped silently hanging — rate-limit, timeout, and SDK errors now surface as in-band error events with the real message. Retest also triggers the completion strip and live feed the same way a full run does.
  • Per-project grader weights: grammar vs. prose × error/warning/info penalties are now six sliders in Settings instead of hardcoded constants. Defaults match previous values, so existing scores are unchanged.
  • Classical plurals bundled in the spellchecker: stelae, phenomena, criteria, corpora, and ~200 others no longer require manual dictionary additions.
  • Per-project category severity overrides: demote noisy categories like TYPOS to "warning" without disabling them entirely.
  • One-click "add to dictionary" button on spelling violations in the benchmark detail view.
  • Etherforce Models tab redesign: new EF Score column with a per-signal breakdown (popularity, cost, quality, speed). The Stats tab merged in — one canonical ranked list for every model.
  • Streaming cost tracking: Lorekeeper, chat, and every other streaming path now records real input/output cost instead of silently reporting zero.

Ambient Awareness

  • Ambient nudges are now a first-class feature — dismissible contextual tips float above the status bar and persist across view and scroll changes.
  • A Session Pickup card on the Desktop sidebar surfaces "where you left off" without digging through recent documents.
  • New Ambient Settings tab with per-rule controls.

Desktop Experience

  • Per-agent ratings in Agent Overview: submit, edit, delete, and dismiss pending rating requests directly from each agent's card.
  • Range sliders across every Settings tab share one consistent native style.
  • Settings forms no longer flag themselves dirty from internal key-order churn, and save resets the dirty state cleanly.
v1.1.7

A faster way to move around the app, smarter Lorekeeper consistency checks, and recoverable document deletion.

Writing & Editing

  • Deleted documents now go to a recoverable Trash list instead of disappearing. Restore or permanently delete whenever you're ready.

Worldbuilding

  • Lorekeeper is scoped to the document it's checking, so consistency findings come from entities that actually appear in the manuscript instead of sweeping the entire Legendry.
  • Lorekeeper scoring is more honest about confidence: low-signal findings are marked as such instead of being weighted the same as strong detections.

Model Integration

  • Clarification now learns from your previous answers, so Ishvana's follow-up questions are sharper and shorter.
  • Etherforce (the model engine) enforces structured output at the provider level — tool calls and analysis results come back in the shape the app expects more reliably.
  • Orchestrator tunables are exposed in Settings so you can dial run behavior per project instead of accepting the defaults.

Desktop Experience

  • Global Ctrl+K command palette: jump to any view, setting, or action without leaving the keyboard. Settings are searchable from inside the palette.
  • Error messages are specific — every failure tells you what actually went wrong instead of showing a generic "something went wrong."
  • Keyboard shortcuts live in a central registry, surfaced through the command palette and a Shortcuts reference page.
v1.1.6

Smarter paste, right-click spell suggestions, and a more resilient local model connection.

Writing & Editing

  • Paste from external sources now keeps structure (bold, italic, links, headings, lists, tables) while stripping visual clutter like font colors and sizes — use Ctrl+Shift+V to keep everything as-is
  • Right-click a spell-scan underline to see suggested corrections, add the word to your dictionary, or ignore it
  • Open documents directly with Ctrl+O or from the File menu
  • New Import tab in the Import/Export dialog with format chips and a file picker

Worldbuilding

  • Clearing optional fields on Legendry entries (summary, excerpt, author) now saves correctly instead of silently keeping the old value

Model Integration

  • Ollama connections fail gracefully when the server isn't running — no more startup crashes or broken model lists
  • Model discovery and embedding initialization survive offline providers without blocking the app

Desktop Experience

  • Autosave serializes writes per document, preventing rare corruption when multiple editor panels share the same file
  • Save-failure messages now tell you what actually went wrong — antivirus locks, disk full, permission issues — instead of a generic error
  • Error reports copied from the error modal now include stack traces for faster troubleshooting
v1.1.5

Sturdier auto-updates and a Legendry toolbar that actually fits.

Desktop Experience

  • Auto-updates are more dependable. Ishvana now keeps a diagnostic log at %APPDATA%/Ishvana/logs/updater.log so stuck updates can finally be traced, and a 20-second safety net kicks off the first update check even if the splash screen or license gate takes an unusual path

Worldbuilding

  • The Legendry browser toolbar now wraps cleanly on narrower panels instead of clipping the sort dropdown and action buttons when the sidebar is collapsed
  • Fixed a regression where the Legendry split-pane resize handle quietly stopped responding to drags — you can now resize the browser column again
v1.1.4

Right-click spellcheck, silent auto-updates, and a steadier model output layer.

Writing & Editing

  • Right-click spell check is now reliable. The custom spelling menu reads fresh context for every click instead of racing against the main process, so the first right-click on a misspelled word shows the real suggestions every time — not last click's stale list
  • Spell replace always targets the word you actually right-clicked. A new click-to-word resolver finds the exact range even when you click slightly off the word or on trailing punctuation
  • Inline highlights from Lorekeeper, Dev Edit, Line Edit, Proofread, and Edit Suggestions no longer end mid-word. Ranges coming back from the backend are now expanded to whole-word boundaries before they're drawn on the page
  • Lorekeeper gained a history view, and its long-running background pass no longer deadlocks on large manuscripts
  • Spell check polish across the editor shell, plus a shared loading indicator so local operations feel instant

Worldbuilding

  • You can now create stub lore entries with no content yet — handy for scaffolding a new faction or location before you have the details
  • Legendry editors and entity tools picked up the same word-boundary fix, so inline flags across the Legendry align with whole words

Model Integration

  • Structured output from LLM providers is now enforced at the provider layer instead of trusting each agent to format itself. Agents that return JSON (outlines, lore entries, analysis passes) hit the right shape on the first try with far fewer retries and parse failures
  • Agents can render Mermaid diagrams end-to-end. A bundled validator sidecar ships with the installer and any agent can call the shared render_diagram tool, so architecture sketches, timelines, and relationship maps generated inside chat are syntactically valid before they render

Desktop Experience

  • The auto-updater now installs silently in the background instead of interrupting you with a modal. Updates download while you work and apply on the next restart
  • Settings has a reset button that restores defaults without you having to hand-edit the config file
  • Accessibility pass across the editor: better focus handling, keyboard paths, and ARIA labeling on the main writing surface
v1.1.3

Minor fixes across chat, writing, timeline, and sales.

Writing & Editing

  • Plot and craft studios now load older projects that have outdated category values. Points with stale labels show up cleanly instead of blocking the whole view
  • Document load carries project and folder context through every step of the pipeline, so documents reopen in the right place after a restart

Chat & Agents

  • Chat history preserves the model's reasoning trace, tool calls, and thinking steps across reloads. Your conversation transcript comes back exactly as it was when you left it
  • Per-message generation stats (speed, context quality, RAG source counts, attached images) now stay attached to the message they belong to after a reload

Worldbuilding

  • Timeline relationship strength and arc point notes show up on the relationship and arc point views instead of being dropped when loading from the database

Sales & Marketing

  • Fixed an XLSX sales import crash. The importer could detect spreadsheets but failed the first time it tried to log one. XLSX imports now work end to end

Under the hood

  • Tightened the connection between the database and every response the UI reads, so columns that drift out of sync no longer slip through as silent 500s
  • Accessibility and visual checks are now part of the end to end suite, catching layout and contrast issues before they ship
  • Hardened chat persistence, sales record round trips, and lore entry loading against bad or legacy data
v1.1.2

Stronger safety net for your manuscript and a denser, sharper Legendry.

Writing & Editing

  • New autorecover writes a hidden .asd snapshot while you work, so an unexpected shutdown leaves your draft recoverable instead of lost
  • Rewritten save pipeline with file locks, a .docx history trail, and a reconciler that catches divergence between disk and the editor before it can bite you
  • Idle main saves quietly commit your work in the background between active edits

Worldbuilding

  • Lore is now called the Legendry across the whole app. The name change carries through menus, panels, and shortcuts to match the wiki and docs
  • Major density pass on Legendry Studio: tighter spacing, more entries visible at a glance, and a cleaner reading rhythm when you're scanning your world
  • Legendry editors have been consolidated into a single workspace with analysis sidebars, so entity, location, and lore editing all share the same shell instead of three different layouts
  • More forgiving date parsing. Ancient or partial years no longer trip up the year extractor when you import older notes

Research

  • Upgraded the embedding store to ChromaDB 1.5.7 for faster and more reliable semantic search across your project
  • Lagan's research queries now degrade gracefully when a collection is empty instead of erroring out, so research works smoothly even on brand-new projects
  • Fixed a bug where filtered semantic searches could return nothing because of an internal include parameter

Desktop Experience

  • Unified left and right rail collapse. Both side rails now respond to the same shortcut and remember their state per workspace
  • Refreshed system tray icon to match the new app identity
v1.1.1

A polished update experience, a fresh look, and steadier lore handling.

Desktop App

  • New in-app update notification — when a release is available, Ishvana tells you inside the app and walks you through downloading and installing it without leaving the editor
  • Refreshed About window with cleaner version info, acknowledgements, and update controls
  • Rebuilt error dialog with clearer messages, copyable details, and direct access to log files

Look & Feel

  • Brand-new Ishvana app icon across the taskbar, tray, installer, and website
  • Cleaner side panels for Conlang Studio, Plot Board, and World Knowledge so they behave consistently with the rest of the app

Worldbuilding Reliability

  • Fixed a race where saving a document with many entities could drop a linked mention in the Legendry
  • More forgiving handling of entity aliases imported from older projects — no more silent alias loss when the stored format is legacy text
v1.1.0

Faster editing workflows, smarter proofreading, and a more polished Legendry. Every module got meaningful improvements. Now available with a 14-day free trial.

Editing

  • Keyboard shortcuts for reviewing suggestions — accept, reject, or skip without touching the mouse
  • Filter suggestions by severity and see pending counts per chapter
  • Your editing progress now saves automatically and survives app restarts
  • Readability scoring (Flesch-Kincaid) with a per-chapter breakdown so you can spot dense sections at a glance
  • Bulk accept or reject suggestions across an entire manuscript with undo

Proofreading

  • Redesigned proofreader with side-by-side view — see suggestions highlighted in context alongside your manuscript
  • Catches more issues: dialogue punctuation, number formatting, contraction consistency, and character name variants pulled from your Legendry
  • Keyboard-driven review — navigate with j/k, pick a suggestion with 19, dismiss with d
  • Proofread an entire book at once with live progress updates
  • Right-click spellcheck corrections now work in all editing views

Legendry

  • Full change history — see who changed what and when, with before/after snapshots for every entry
  • Select multiple entries at once to move, reclassify, or delete in bulk
  • Breadcrumb navigation lets you click through Library > Category > Type to filter quickly
  • Your preferred view, sort order, and filters are remembered per project
  • Each entity type (character, location, world, etc.) now has its own tailored detail layout
  • Editing a lore entry immediately checks for contradictions instead of waiting for the overnight scan
  • Data recovery tools if your search index ever gets out of sync

ProseGuard

  • Suggest Rules — analyzes your manuscript and recommends rules based on your POV, character names, and writing habits
  • Rules you haven't configured yet now show up with setup hints so you can discover what ProseGuard can do
  • Fewer false positives thanks to smarter word-boundary matching

Councils & Outline

  • Accepting Council drafts into your Legendry now supports undo — one click to revert if you change your mind
  • Review and re-run past Council sessions from your history
  • Select and update multiple outline nodes at once with the bulk toolbar

Desktop App

  • The app recovers automatically if something goes wrong — detects crashes, restarts the backend, and shows a status banner
  • Undo notifications are less noisy and show clearer descriptions of what changed
  • Your world's naming rules and constraints are now respected by all agents during generation
  • 14-day free trial — try everything before you buy, with a countdown banner so you always know where you stand
  • Installer is now code-signed by Swordsfall Studios LLC — no more "Unknown Publisher" warnings
  • Models are bundled with the installer — no downloads needed on first launch
  • Improved stability and error reporting throughout
v1.0.0 Initial Release

First packaged release of Ishvana. Internal build for testing the installer pipeline, auto-updater, and license verification system.

Editor & Workspace

  • TipTap-powered rich text editor with 27 extensions, slash commands, and @mentions
  • Auto-save with full revision tracking
  • Manuscript-first desktop surface with context sidebar (lore, knowledge, prose rules, agents) adjacent to the page
  • Revision pass workflow: drafting, continuity checks, prose cleanup, and publish prep in the same environment

Legendry | Worldbuilding Database

  • 12 specialized entry types: Characters, Locations, Factions, Items, Events, Concepts, Species, Languages, Religions, Legends, Articles, and Custom types with user-defined fields
  • 18 relationship types (parent, sibling, spouse, ally, enemy, mentor, member, leader, creator, destroyer, and more) with automatic bidirectional creation
  • Interactive force-directed d3 relationship graph with type filtering, zoom, pan, and node pinning
  • Hybrid search: keyword + semantic combined. Query by meaning ("the faction that controls trade in the northern provinces")
  • Voice profiles per character: dialect, vocabulary level, speech patterns, verbal tics
  • Three-tier visibility: Known / Unknown / Secret per faction and character
  • Gallery, list, and card views with filtering by type, category, tag, or custom criteria
  • Automated lore ingestion: paste raw worldbuilding notes, Ishvana extracts structured entries into the correct types

Maps | Interactive Cartography

  • Accepts PNG, JPG, WebP, and SVG up to 100MB per canvas
  • Nondestructive image swapping: all coordinates, pins, polygons, and visibility layers mathematically anchor to the new image
  • Drill-down parent hierarchy: Continent → Province → City → Inn Interior with breadcrumb navigation
  • 4 annotation primitives: Pins (Phosphor icons, 3 sizes, GM notes, hover tooltips), Polygon Regions (unlimited vertices, fill/stroke control), Polyline Paths (5 stroke styles, midpoint arrows), Typography Labels (360° rotation, letter-spacing, italicization)
  • True scale calibration (miles, km, leagues, days-travel) with Measure Mode (M) for distance calculation
  • Per-object zoom visibility thresholds and dynamic legend control with color-coded categories
  • Keyboard-first controls: V/P/R/A/T/M/Esc/0/+/- with smooth-scroll zoom (-3 to +5, 0.25 increments)
  • 6-second soft-undo intercept on accidental map deletions

Character Knowledge | Continuity Tracking

  • Knowledge matrix: every character against every piece of information, color-coded (known, partial, unknown, secret)
  • Per-character fact tracking with learned-at timestamps (scene, chapter, method of discovery)
  • Deterministic scanning: rule-based violation detection. Same input always produces same output
  • Character Council panel: personality, relationships, and knowledge state live beside the editor
  • Semantic violation detection: implicit knowledge violations, emotional bleeding, transitive knowledge inference

ProseGuard | Deterministic Prose Linting

  • 5 rule categories: Structural (sentence length, pacing), Stylistic (adverbs, passive voice, purple prose), Semantic (cliches, mixed metaphors), Continuity (name consistency, timeline violations), Machine Tells (hedging, formulaic transitions)
  • 4-level scope hierarchy: Project → Document → Scene → Character
  • 4 severity levels: error, warning, info, hint
  • Rule editor with live previews. YAML import/export for sharing and version control
  • Deterministic enforcement: your rules, applied identically every time

Mechanics | Rules Engine

  • Define complete mechanical systems: stats, formulas, dice notation, computed fields
  • Dice expression support: 2d6+4, 1d20+STR, 4d6kh3 with real-time validation
  • Probability analysis: pool distributions, opposed rolls, power curves, DC tier breakdowns
  • 3 preset systems: D&D 5e (ability scores, proficiency, AC, spell slots), Pathfinder 2e (four-tier proficiency, three-action economy), Fate Core (Fate ladder, aspects, stress tracks)
  • Rules Lawyer: stat validation, formula consistency checking, rule conflict detection

Lorekeeper & Hawken

  • Lorekeeper consistency scoring (0–100), contradiction detection, issue lifecycle (open → acknowledged → resolved → dismissed)
  • 3 analysis depths: quick scan, standard cross-reference, deep semantic
  • ML pattern analysis: topic clustering (TF-IDF), anomaly detection (Isolation Forest), coverage mapping, maturity scoring
  • Hawken writing coach: lore-aware RAG pipeline (Query Legendry → Load voice profiles → Check ProseGuard rules → Generate)
  • 14 writing styles: Literary, Commercial, Poetic, Minimalist, Maximalist, Dark, Humorous, Academic, Journalistic, Epistolary, Stream of Consciousness, Hard-Boiled, Gothic, Mythic
  • Hawken Edit: inline transforms (tighten, voice match, tone shift, style transform) that preserve world vocabulary and character voices
  • ProseGuard integration: active rules sent as pre-generation constraints

Studio Tools

  • Visual Studio: board-based asset management for reference art, portraits, and media linked to lore entries
  • Publish: compile to PDF with platform presets (Amazon KDP, IngramSpark), edition control, batch export
  • Analytics: writing streaks, session tracking, goals, model benchmarks
  • Research: built-in browser with smart bookmarks, entity extraction, project relevance scoring
  • Marketing: genre analysis, comp title tracking, reader persona development, KDP metadata

Infrastructure & Packaging

  • Windows NSIS installer with desktop and Start Menu shortcuts
  • Auto-updater checks for new versions on startup
  • License activation with offline verification
  • Code-signed via Azure Trusted Signing
  • Bundled: Python 3.12 backend (FastAPI), JRE 21, LanguageTool, Typst, Libertinus fonts
  • Electron desktop shell with Vue 3 renderer
  • SQLite + ChromaDB (vector embeddings) for local-first data
  • All data stays on your machine. No cloud dependency.

What's Next

See v1.1.0 above for what shipped since this initial build.

Technical Specifications

PlatformWindows 10/11 (64-bit)
Min Display1920 × 1080
Min RAM16 GB
Min CPU6-core / 12-thread
GPU (for local models)Nvidia 8 GB VRAM
Backend Port37737 (localhost)
DatabaseSQLite + ChromaDB
License$99 one-time, 3 devices

Get the latest build now.

$99 one-time. Every feature included. 14-day money-back guarantee.

Purchase Ishvana