Desk 01
Timeline
Character bible, scene, and props on one board; the 6 / 10s clip track stays underneath.
YSK product · AI short-drama studio · MIT · v1.11.4
InstantDrama Magician
Lock a cast. Board every beat. Then work the same story three ways: a
timeline of 6 / 10 second clips, a comic book of full pages you can still turn into film, or a
key-art desk for covers, stills, promo frames, and hero headshots.
Electron desktop + browser remote + CLI instant-drama (190 channels).
v1.11.0 adds a spatial shot-stage and an optional Blender bridge for white-model / clay playblast I2V (Blender is not embedded).
Desk 02
Comics
Lock the printed page, then animate it or shoot it as a short-drama clip. Export does not rewrite the timeline.
Desk 03
Key art
Covers, stills, promo, headshots. Four make methods, then the take. Cover is never automatic.
Three desksTimeline · Comics · Key art
190CLI channels
Chapters firstWrite the story, then suggest from plot
FFmpegExport, xfade, subtitles
From idea to finished film — three ways
Sidebar: Stories · Characters · Costumes · Scenes · Props · Actions · Comics · Key art · Timeline · Activity · Settings. Write chapters first, then Suggest from story when filling libraries.
- Story / chaptersWrite the bible, then split filmable beats
- Asset librariesCharacters / costumes / scenes / props / actions
- Pick a deskTimeline, comics, or key art
- GenerateStills, comic pages, or publicity → video
- ExportFFmpeg; comics can ship its own film
Timeline desk
Linear 6 / 10s clips, snap and pack, advanced prep for continuity, then FFmpeg export.
Comics studio
Even grids or manga panels, 9:16 / 1:1 / 16:9. Lock the printed page, then animate it or shoot a short-drama clip.
Key-art desk
Eight publicity types, four make methods (new / edit / lock face / continue). Setting the story cover is never automatic.
Spatial shot-stage
In-app spatial shot-stage for blocking and camera; optional Blender add-on for white-model / clay playblast I2V. InstantDrama does not embed Blender.
Remote + CLI
Browser remote on the same data; 190 channels for scripts, CI, and agents.
| Area | What you can do |
|---|---|
| Stories | Multi-story management, cover AI, style bible (advanced style / hard-rule kits, apply-only), chapters tab, cast binding, .idm.zip backup |
| Chapters / beats | Write chapters first; split plot beats. Per beat: characters max 4, scenes max 2, props max 4, actions max 4 |
| Characters | Global cast, soul.md / SoulMD Hub, multi-angle sheets, identity lock, advanced field kits (appearance / costume / voice / mannerisms / hard rules), photo book, vision fill, Suggest from story |
| Costumes | Wardrobe library, try-on dual-write, advanced look / hard-rule kits; vision fill or Suggest from story |
| Scenes | Copy, plates / looks / atmosphere, advanced location / set-dressing / camera / hard-rule kits, gallery, vision AI or Suggest from story |
| Props | Descriptions, advanced look / hard-rule kits, master prompt, plate variants, timeline binding, vision fill or Suggest from story |
| Actions | Multi-panel boards (2–6), advanced motion / camera / hard-rule kits, cast refs, vision fill or Suggest from story |
| Comics | Even + manga layouts, 9:16 / 1:1 / 16:9, page animation or short-drama shot, versioned page videos, export pages that already have video (does not rewrite the timeline) |
| Key art | Eight types, four make methods, versioned PNGs, optional set as story cover |
| MediaGen | Materials → multi-image vision polish → image / video; recipe picker |
| Gallery UI | EntityGalleryPanel: preview, cover, identity multi-select, intro video |
| Timeline | Linear layout, snap/pack, 6s/10s clips, dialogue & camera tags; pipeline board stacks to the live window height |
| Advanced prep | Cast lock → storyboard stills (end-frame continuity) → video; per-cell refine |
| Audio / subs | TTS mix, burn-in subs, xfade / ducking, aspect-aware export |
| Settings | LLM / image / video providers, OS completion notifications, light / dark / system, FFmpeg, web server, auto-update |
| Spatial shot-stage | In-app blocking and camera stage; optional Blender add-on for white-model / clay playblast I2V. InstantDrama does not embed Blender |
| CLI / remote | 190 channels (spatial package adds five); in-app web or standalone instant-drama server |
| i18n / updates | 10 UI languages; PromptCatalog + packs native; packaged GitHub Releases auto-update; CLI instant-drama update |
Creative asset libraries
Reusable global assets — build once, share across stories. Write chapters first; Characters / Costumes / Scenes / Props / Actions can Suggest from story.
Stories
- Create / edit / delete independent short-drama projects; status, cover, sort
- Cover zoom / regenerate / save as; import/export story
.idm.zip - Basics: cover, AI quick create (style note / beats), title, status, art style, style bible (must / must-not; advanced style / hard-rule kits apply-only)
- Chapters: write narrative bodies (AI fill / polish); this is the plot source for later fills
- Cast: link characters, scenes, props, actions (search + in/out filters)
Chapters & plot beats
- Chapters first: narrative bodies become the plot source for later fills; AI can fill or polish chapters
- Suggest from story (Characters, Scenes, Props, Actions, Costumes, beside AI fill): check chapters (
chapter:<id>) and beats (beat:<id>). Desktop default-checks beats already bound to this asset; empty selection = whole story (chapters first) - Plot beats: split chapters into filmable beats; per beat characters max 4 / scenes max 2 / props max 4 / actions max 4; beat screenplay text
Characters
- Global cast library; filters for gender, art style, image, soul, language
- Profile: name, description, age, gender, language, voice; advanced field kits on appearance / costume / voice / mannerisms / hard rules; AI fill from concept/draft/soul, still-only vision, or Suggest from story
- References: multi-angle bible, body/base/costume pipeline, external refs, Lock identity, intro video with shot templates
- Reference schemes: turnaround packs, face locks, costume plates — pick a scheme, then generate
- Photo book: albums of scene / prop / action stills, then slideshow or AI clips (
characters:renderPhotoBook) - Costume tab binding; SoulMD Hub index + soul.md import
Costumes
- Global wardrobe linked to 0…N characters
- Advanced look / hard-rule kits apply into the description fields (apply-only; no kit bag on reopen)
- Vision fill from reference stills only, or Suggest from story; multi-image gallery, cover, intro video
- Identity multi-select thumbs; try-on with character refs
- Accept try-on draft dual-writes to character and costume galleries
Scenes
- Description and script fields; advanced kits on space / set dressing / camera / hard rules; plates, looks, atmosphere
- Establishing, hero, interior, weather — swap atmosphere on the same set
- Scene gallery and variants; vision AI fill from selected/cover still, or Suggest from story
Props
- Name, description; advanced look / hard-rule kits; master prompt, plate variants (hero, three-quarter, detail, material, scale)
- Timeline clip binding; vision AI fill from reference stills, or Suggest from story
Actions (motion guides)
- Global action library for body / blocking / staging cues
- Advanced kits on tempo & body / camera / hard rules
- Multi-panel boards: 2–6 panels (strip or 2×2 / 2×3); panel 1 = first beat, panel N = last
- Art style, external refs, cast refs from characters/costumes/scenes/props
- Vision AI fill, or Suggest from story; accumulate panels, reorder, cover
- Attach to story cast, beats, and timeline clips; inject pacing / intent / camera notes; optional i2v reference
MediaGen shell
- Unified: pick stills and notes → multi-image vision polish → image / video
mediaGen:extract·polish·generateImage- Shared by library pages and timeline refine (timeline-still / timeline-clip)
- Recipe picker before LLM improve/generate; chrome localized
Gallery UI
- Shared EntityGalleryPanel: large preview, zoom / save-as / cover / remove / intro video
- Thumbnail row: preview vs identity multi-select
- Labels and errors localized
Three production desks
Same story: timeline film, comic pages, key-art publicity. Comics export does not rewrite the timeline.
Timeline
- Play / Undo / Redo; batch Generate; Export + history
- Duration, ready count; AI clips 6s or 10s only
- Zoom, timeline snap, snap grid, pack clips
- Clip editor: bind characters / scenes / props / actions, duration, beat screenplay tags
- Per clip generate / regenerate / continue; cancel; retry-failed-only
- Pipeline board stacks to the live window height: character → scene → props → action, track underneath
- Export options: TTS, burn-in subs, xfade, ducking, aspect ratio
Comics studio
- One comic book per story; pages are full multi-panel stills, not the timeline
- Layout: even grids + manga irregular panels; format tall 9:16, square 1:1, wide 16:9
- Panels: caption, bind a timeline beat, art style (inherit the book or override)
- Artwork then two video schemes: page animation (lock gutters and print layout, gentle camera) or short-drama shot (same lock / camera / materials polish as timeline Generate video)
- Versioned page videos in a grid; set primary for export; delete only when you choose
- Export film concatenates pages that already have video (
clipSource: comics) — does not rewrite the story timeline - Optional import of page stills onto the timeline as continuity locks
Key-art desk
- One key-art book per story; each item is a publicity still (not a comic page, not a timeline clip)
- Eight types: cover poster, tall poster, production still, promo, social story, hero headshot, bust, cast lineup
- Format 9:16 / 1:1 / 16:9 follows the type, or you override
- Materials: brief, art style, multi-select characters, scene, timeline beat, optional comic page
- Four make methods: new image (refs lock identity only); edit this still (this take is the pixel base); lock the face, redraw (same person, new pose / crop); continue last still (same people and place, next beat)
- Versioned stills in a grid; set primary; delete only when you choose
- Set as story cover writes the cover path — never automatic
Advanced prep
- From Timeline → Advanced
- Cast lock on-screen looks
- Storyboard stills: tail-frame continuity, prev keyframe base, multi-ref polish, backfill missing, status labels
- Video queue when stills ready (skip existing optional)
- Per-cell refine still or video via MediaGen
Audio / subtitles
- Optional TTS mix and burn-in dialogue subtitles
- xfade transitions, BGM ducking, aspect-aware export
Tools & settings
Activity log, providers, completion notifications, i18n, and auto-update.
Activity log
- Local activity.jsonl-style events: generation, export, update, support reports
- Debug API / pipeline issues
Settings
- LLM: OpenAI-compatible; default Grok Gateway (:3847); OpenAI / Custom / Kimi (Moonshot)…
- Image: shared or independent base URL / key / model (e.g. Seedream)
- Video: auto / http / stub; Seedance, Grok
/v1/videos; 6/10s; poll & timeout - Diagnostics, FFmpeg path, web server, auto-update, support report (keys redacted), legal AUP
- OS completion notifications when generation finishes
- Appearance: light / dark / system
i18n & auto-update
- zh-HK, zh-CN, en, es, hi, ar (RTL), pt-BR, fr, ja, ru
- Recipe picker before LLM improve/generate; MediaGen chrome localized; PromptCatalog + packs native (no English paste)
- Packaged electron-updater against GitHub Releases
Honest limits: final look depends on your models and prompts. This tool owns workflow, continuity, and export — not guaranteed cinema-grade auto film. Store signing / notarize needs your certificates.
CLI & web remote
Headless scripts, agent integration, or browser control of the same project data.
CLI instant-drama
- Requires Node.js 20+;
npm i -g instant-drama-magician - local vs remote (
/api/invoke+ Bearer) - 190 channels covering desktop IPC; OpenClaw / Hermes agents
- Domains: stories, chapters, characters, costumes, comics, keyArt, mediaGen, timeline, videoPrep, actions, generation, server, spatial, tools schema…
- e.g.
instant-drama chapters list;scenes ai-fillacceptssuggestFromStoryandsegmentKeys instant-drama update/ pin a version;doctoralso reports npm update status
Web remote & self-host
- A: in-app web server (Settings) — same userData as desktop
- B: standalone
instant-drama serverwith data dir + auth token
Screenshots & install
v1.11.4 desktop UI. Filter shots by workspace, then follow install and the recommended flow.
01
Story management
Multi-project list, covers, status, asset counts, search, import/export backups.
02
Story editor (Basics)
Cover, AI style note, title, art style, style bible (must / must-not).
03
Story chapters
Write the narrative first. AI can fill or polish; later fills use chapters as the plot source.
04
Plot beats
Split chapters into filmable beats. Bind character, scene, prop, and action on each beat.
05
Bind cast, sets, props
Browse the global libraries and attach what this story actually uses.
06
Character library
Global cast, multi-image cards; gender / style / soul / language filters.
07
Character bible
Identity, body, costume, and detail stills plus the written profile and hard rules.
08
Reference schemes
Turnaround packs, face locks, costume plates — pick a scheme, then generate.
09
Costume try-on
Pick the wearer, a body-plate, and a pose. The still writes to both costume and character galleries.
10
Scene plates
Establishing, hero, interior, weather — generate or swap atmosphere on the same set.
11
Prop stills
Hero product, three-quarter, detail, material, scale.
12
Comics layouts
Even grids and irregular manga panels. Format (9:16 / 1:1 / 16:9) follows the template.
13
Finished page
Two video schemes: animate the printed page, or shoot it as a short-drama clip.
14
Key-art types
Cover, poster, still, promo, social, headshot, bust, lineup.
15
Make methods
New image, edit this still, lock the face, continue the last take. One click sets the story cover.
16
MediaGen clip prep
Previous still, this beat’s still, character, scene, then lock continuity in text.
17
Timeline track
Snap, 6s / 10s clips, bound assets, beat screenplay, per-clip generate.
18
Timeline pipeline
The same beats as a board: character → scene → props → action, with the track underneath.
19
Advanced prep
Cast lock → storyboard stills (continuity) → video. Continuity lock and per-cell refine.
20
Languages & notify
Ten UI languages, light / dark / system, OS completion notifications.
21
Providers
Local Grok Gateway by default, plus cloud and local LLM / image / video cards.
Recommended flow
- Settings → API key → Test Chat
- Stories → chapters (narrative) → script beats
- Characters → multi-angle sheet → lock identity (or still-only vision fill, or Suggest from story → pick chapters / beats)
- Scenes / Props / Costumes / Actions assets
- Story Cast → link assets including actions
- Chapters and script beats → bind assets per beat
- Timeline → layout clips, beat screenplay
or Comics → pick a template → write panels → generate the page → page video
or Key art → pick a type → bind cast → choose a make method → generate stills / set cover - Advanced prep → stills (continuity) → video
- Export (optional TTS / subs); Comics can export its own film from pages that have video
npm install -g instant-drama-magician
instant-drama doctor --json
instant-drama --local stories list --json
instant-drama chapters list --args '["STORY_ID"]' --json
instant-drama channels list --json # ~190
npm install && npx prisma db push && npm run devOpen source to use. Custom builds for enterprises.
InstantDrama Magician is MIT-licensed. YSK Limited can custom-develop on that base for companies — no charge for using the open-source product; custom work is quoted in writing by scope.
Open source
Self-host, evaluate internally, or contribute.
- Desktop, CLI, and browser remote
- GitHub Releases and npm, install yourself
- No fee to use the open-source product
YSK custom build
Production lines, brand rules, private networks, ongoing change.
- Extend instant-drama-magician as the base
- Hong Kong company contract; NDA if needed
- No custom prices on this page — written quote only
Studios & short-drama lines
Shared templates, review steps, and character bibles across many projects.
Brands & IP
White-label UI, default style bibles, and must / must-not compliance.
Agencies
Client-specific desktop flows, wired to existing libraries or internal accounts.
Teaching
Classroom installers, demo stories, and remote training.
| Custom scope | What we can do |
|---|---|
| Brand / white-label | UI, installer name, default style bible and asset libraries |
| Workflow | Extra desks, review, brand rules, production templates |
| Deploy | On-network instant-drama server, customer hosts, installers |
| Store signing | We configure signing / notarize; you supply certificates. We do not apply for Apple / Microsoft accounts |
| Providers & models | Your OpenAI-compatible gateway, gctoac, or private LLM / image / video (data-stays-in-region as a separate AI engagement) |
| Integrations | Existing DAM, CMS, media libraries, internal accounts |
| After delivery | Docs, remote training, warranty window; then maintenance by agreement. Small changes can use outsourcing monthly tickets |
How an engagement runs
From first contact to warranty, every step has a written scope. Open-source use and custom work are billed separately.
- EnquireWhatsApp or email: studio line, brand content, teaching, or IP
- DiscoverRemote call: current flow, providers, compliance, private models
- ProposeScope, timeline, deliverables, signing / LAN / private models; written quote
- ContractHong Kong company agreement; NDA if needed
- BuildExtend instant-drama-magician; staged demos
- AcceptInstaller or LAN deploy, training, docs
- WarrantyFixes in the agreed window; then maintenance or monthly outsourcing
Deliverables (per contract)
- Custom desktop installer and/or in-network server
- Deploy and operations docs; remote training
- Defect fixes in the agreed warranty window
- Upstream core stays MIT; custom-module IP is defined in the contract
Honest limits: look still depends on your models and prompts — a custom build does not guarantee cinema-grade auto film. This is a desktop / local data-root product; YSK does not offer a multi-tenant SaaS InstantDrama.
FAQ
Yes. MIT-licensed desktop, CLI, and browser remote. YSK does not charge for using the open-source product. Custom work is quoted by scope.
White-label, LAN, signing, private models, and integrations vary widely. Quotes are written. Small changes can use existing outsourcing monthly tickets; large work is a project.
The open-source core stays MIT. Use and ownership of custom modules are set in the contract. This page does not assume all source goes to the client or a perpetual exclusive.
Yes as custom scope: we configure signing and notarize; you supply certificates. We do not apply for Apple or Microsoft developer accounts.
Get started with InstantDrama Magician
MIT open source to download yourself; YSK Limited can custom-build a production line on top.