ARIS-Movie-Director
Turn fuzzy story ideas into cross-model-audited comic movies
All Skills in This Repository (16)
Pure Emerald Level Indicatorscomic-continuity-audit
Audit comic panels against motif state tables and global invariants.
comic-blind-comparison-review
Compare two complete comics via double-blind A/B review using Codex and Gemini models.
method-figure
Generate publication-grade method figures from a deterministic JSON blueprint.
comic-intent-parser
Convert raw user briefs into schema-valid intent_spec nodes for comic creation.
comic-outline-creator
Transform a locked comic skeleton into an approved outline with a 6-column beat table.
movie-pipeline
Generate image-based movies from story ideas with automated review loops.
comic-json-compiler
Compile comic.json from locked storyboard and blueprint data with schema validation.
comic-director
Converts authored comic.json narratives into pixel-art frames with 3-reviewer cross-model verification.
comic-blueprint-author
Generate deterministic SVG blueprints for comic panels from locked specifications.
comic-asset-ref-generator
Generate canonical identity references and deterministic SVG motifs for comic assets.
comic-storyboard-creator
Generate a comic storyboard with fixed page order and continuity ledger.
comic-panel-prompt-builder
Compose deterministic image generation prompts from comic panel specifications.
Frequently Asked Questions
FAQPage SchemaHow to install ARIS-Movie-Director?โผ
Run `npx skills add wanshuiyin/ARIS-Movie-Director --all -g -y` in your terminal to install all skills in this suite globally.
What does ARIS-Movie-Director actually produce?โผ
It turns a fuzzy story idea into a multi-scene, image-based comic movie with a single-file HTML viewer. Every panel is baked by an image model and verified by independent reviewers before it ships.
How does it stop generated frames from changing facts?โผ
It locks expected numbers, labels, and character identity into a blueprint before rendering, then blind-transcribes each baked frame with different model families and runs a deterministic token diff. A beautiful frame with a wrong number is rejected.
Which AI agents and models does it work with?โผ
The skills follow the standard SKILL.md format and run in coding agents like Claude Code and Codex. The pipeline uses Codex for image baking and Gemini plus Codex as independent reviewers.
Can it generate video instead of still images?โผ
Not yet. This release produces image-based movies told in baked still frames; video-based generation is the planned next step.
Related Repositories in Design & Creative
View All in Design & Creativeโui-ux-pro-max-skill
AI design intelligence for professional UI, brands, and presentations
ui
Build and customize beautiful React component libraries with AI
open-design
Open-source AI design studio for decks, prototypes, images, and video