@docyrus/animaguru 0.16.0 → 0.16.2

package/skills/animaguru-art-motion/references/sources.md

@@ -0,0 +1,40 @@
+# Sources, translation and scope
+
+This skill is an English AnimaGuru adaptation of [Huashu Art Motion](https://github.com/alchaincyf/huashu-art-motion), by alchaincyf (Huashu), discovered through [the Voxyz_ai post](https://x.com/Voxyz_ai/status/2109690569563746755).
+
+- Pinned source revision: `2571ee0a9a0467c4d99e935ea9e0fadda19210b0`.
+- Reviewed on October 15, 2026.
+- Upstream code and documentation: MIT, with the original copyright and license retained in [LICENSE](../LICENSE).
+- Coverage: all 36 art recipe cards and all nine explainer grammar cards, mapped individually in [coverage.json](coverage.json).
+
+The post described 35 art styles. At the pinned revision the repository has 36: its latest standalone SVG cartoon card has source ID `37_rick_morty`, while ID 07 has no card. The older 35-scene gallery and the SVG example use different render paths. Nine grammars are documented; eight have parameterized upstream examples, while the presenter card has a reference film. This skill bundles English reference guidance, not those engines, examples, characters or media.
+
+## What was translated and adapted
+
+The [art index](art-styles.md) links English cards for every source recipe. Each retains material construction, useful palette/parameter examples, motif movement, transition ideas and known limitations. The [grammar index](explainer-grammars.md) covers every source grammar's purpose, composition, choreography, camera, timing and common failures. Individual cards link back to the pinned upstream document.
+
+Shared methods from the source's analysis, mechanism, still-first routes, code-rendering guide, rhythm/audio, lessons, author contract, grammar integration, character, long-scroll, narration and camera documents are consolidated in [production](production.md) and [rendering](rendering.md). This is a production translation and adaptation, not a line-for-line archive of every demo log or private review conversation.
+
+English copy is fitted by real width, voice and reading time. Chinese glyph-count and typing-speed examples do not become English limits. Names, source IDs and original copyright attribution are preserved where they identify provenance. Source numerical values are marked as starting points from upstream studies; historical quality ratings and QA tables are not presented as current AnimaGuru test results.
+
+## Deliberate AnimaGuru changes
+
+Use the installed AnimaGuru generation tools, project assets, enabled models, cue sheet, PixiJS kit, player, audio tools and exporter. The upstream Canvas global APIs and author-specific speech endpoints are not installed or assumed. Generated layers, procedural art and hybrid production are all available decisions; advanced character quality should be judged from the actual result.
+
+The reusable guidance does not import upstream promotional instructions, update checks, model-brand rankings, fixed demo coordinates, compulsory three-direction reviews, blanket character bans, hue bans, fixed motion percentages or automatic delegation. Those reflected the original author's environment or preferences. It preserves the evidence behind useful lessons: material-specific motion, deliberate cameras, stable random variation, correct sources and numbers, visible contact and honest review.
+
+## Asset exceptions
+
+The upstream README separates some assets from MIT code/documentation:
+
+- Arphic-licensed stroke median data has its own terms.
+- Bundled fonts have their own licenses, including OFL 1.1.
+- The author's likeness, character sheets and footage containing that likeness are demonstration material.
+- The Clawd mascot belongs to Anthropic; fan-demo footage is not general MIT media.
+- Some museum artwork is separately designated CC0.
+
+No such assets, voices, fonts, stroke data, footage or mascots are redistributed with this skill. The MIT notice does not authorize their reuse. Source styles referencing artists, studios or programs identify influences; use original or authorized subjects. When adding assets later, check their individual terms and preserve their notices.
+
+## Extending the references
+
+Record the exact source revision and distinguish an observed result from a hypothesis. Keep project-specific private evidence in that project's production records. A stable general lesson can update the relevant card; correct an obsolete sentence instead of accumulating contradictory rules. If copying substantial new source material, preserve applicable attribution and confirm asset exceptions again. The temporary analysis checkout is not a runtime dependency.

package/skills/animaguru-cutcraft/SKILL.md

@@ -0,0 +1,48 @@
+---
+name: animaguru-cutcraft
+description: Create AnimaGuru Cutcraft explainers and product stories with tactile paper collage, precise editorial typography, purposeful motion, and AnimaGuru-generated media. Use for this visual treatment or a Vox-inspired explainer adapted to AnimaGuru.
+---
+
+# AnimaGuru Cutcraft
+
+Make an original editorial film that explains an idea through things the viewer can see move, connect, separate and resolve. Use AnimaGuru for generation and a controlled composition for titles, labels, timing and branding. The style can serve another product; use that product's facts and identity.
+
+## Establish the film
+
+Reuse the user's subject, references, language, placement and generation authorization. Resolve only consequential missing choices. For an exploratory sample, state sensible assumptions and proceed: a short English landscape film with a single narrator is a useful starting point, not a requirement for every film.
+
+Read first-party product material before writing claims. Separate actual product footage from conceptual illustrations in the production records. Avoid unverified performance claims, invented customer evidence, and promises of perfect AI results. End an advertisement with a relevant invitation; end an educational explainer on its insight unless a CTA was requested.
+
+Write a concise narration and a beat sheet pairing each spoken idea with a visible event. Duration follows the measured voice and the user's brief. A problem, a concrete mechanism, an outcome and a short close can work for a product film; choose the structure that fits the subject.
+
+## The Cutcraft visual language
+
+- **Material:** matte colored cardstock, visible fibers, slightly irregular scissor edges and short directional contact shadows. Layer flat pieces; keep silhouettes readable. Avoid glossy plastic and unrelated photorealistic environments.
+- **AnimaGuru palette:** warm cream `#F7F2E8`, ink `#162B38`, brand blue `#2C6BD0`, orange `#E67D43`, with restrained sage `#93AA8B`. For client films, replace accents with the client's brand. Backgrounds should have quiet areas for type.
+- **Typography:** one bold editorial heading family and a compatible reading face. Use actual local fonts, vector logos and rendered text for names, numbers and claims. Keep generated footage free of essential lettering. Short labels summarize; narration carries the full sentence.
+- **Composition:** one dominant idea per beat, useful negative space, visibly separated foreground and background. A print-like grid, small folio marks and a fine blue connector line give the film its own identity. Avoid copying Vox's logo, layouts or other identifying brand elements.
+- **Motion:** cutouts slide, rotate at plausible paper joints, pin into place, peel, stack or reveal. A connector is drawn when a relationship becomes clear. Motion answers the spoken line; texture and camera movement remain restrained.
+- **Continuity:** lock material, light direction, palette, camera language and recurring character description. An orange fox with a cobalt scarf is an optional AnimaGuru motif, not compulsory casting for other subjects.
+- **Sound:** one narrator across the film, a quiet instrumental bed and a few motivated paper taps or rustles. Keep speech intelligible and let the closing thought breathe.
+
+Build a visual explanation instead of using the same decorative collage under every sentence. If illustrating control, visibly move one object while its surroundings stay fixed. If explaining a connected workflow, show the pieces joining in the order the narration names them.
+
+## Generate with AnimaGuru
+
+Read [the generation adapter](references/generation.md) when producing media. It covers CLI discovery, supported references, credit estimates, continuity and local assembly. Do not substitute MaxFusion calls or assume a Google Omni model exists.
+
+Prefer the installed AnimaGuru CLI for local asset work; use connected MCP tools when they are available and appropriate. Creating this skill is not generation authorization. A user request to generate the film authorizes a proportionate initial attempt; honor stated budgets and surface a concrete estimate before submission. Do not purchase credits or change provider defaults as part of film production.
+
+Generate artwork and motion separately from exact type. For a short trial, a few strong compositions with changing editorial labels usually communicate more than many loosely related clips. Reuse good existing textures, music or SFX when they fit. A generated clip may be one layer inside a controlled film rather than the entire frame.
+
+For precise placement or when video providers are unavailable, generate isolated transparent paper subjects and animate their transforms in the composition. Build the fixed setting from separate paper layers. This is a complete production route: character movement, path reveals, clip placement and typography can remain editable without a video-model call. Describe it as animation of generated artwork, not as AI-generated video footage.
+
+Use a single speech asset or consistently configured segments. Measure the recording before finalizing scenes; do not assume the requested pacing was achieved. Replace or mute incidental speech from video generation when using the separate narrator.
+
+## Review and deliver
+
+Keep the brief, narration, scene timings, request settings, estimates, job IDs, returned asset IDs, costs and source links together with the film's production files. Preserve an editable composition or reproducible render recipe and local copies of the media.
+
+Inspect the actual opening, middle, transitions and closing frames. Check exact brand spelling, readable labels, character consistency, contrast, crop and any model artifacts. Check audio duration and clipping; listen when playback tools permit, and describe listening limitations honestly. Repair the composition before buying another generation when timing, crop or typography can solve the problem.
+
+Export a playable local film and report what was generated, what was reused, and any remaining defects. Return a preview, the final file, and a link to the skill. Creating a film does not imply publishing it publicly or changing the product website.

package/skills/animaguru-cutcraft/agents/openai.yaml

@@ -0,0 +1,4 @@
+interface:
+  display_name: "AnimaGuru Cutcraft"
+  short_description: "Paper collage explainers made with AnimaGuru"
+  default_prompt: "Use $animaguru-cutcraft to make a short AnimaGuru explainer with tactile paper collage and precise editorial motion."

package/skills/animaguru-cutcraft/references/generation.md

@@ -0,0 +1,61 @@
+# AnimaGuru generation adapter
+
+## Discover the actual environment
+
+For CLI work, inspect `animaguru --help`, `animaguru schema`, `animaguru auth status`, `animaguru credits` and `animaguru catalog`. Do not read credential files. If it is absent, use connected MCP or obtain agreement to install the CLI; installation and authentication are separate actions.
+
+List projects before choosing one, then inspect the selected project and `assets list PROJECT_ID`. Reuse the user's project where appropriate. Use returned project IDs, asset IDs and file SHAs. Create a separate project when the film needs its own asset container. Saved filenames are not file SHAs.
+
+The CLI catalog describes known capabilities; the live project's `availableModels` and `videoProfiles` describe enabled models. Choose from their intersection. A local `--dry-run` can succeed for a disabled server model, so it does not prove provider availability.
+
+With MCP, discover the equivalent project, catalog, assets, estimate, generate and job tools. If account linking is needed, let the user complete it. Never invent tool calls or fall back to database access.
+
+## Plan and estimate
+
+Save requests as JSON and validate each with `assets generate ... --dry-run` or a batch with `assets batch PROJECT_ID --file REQUESTS.json --dry-run`. Add all estimates, including speech and any music/SFX, before the first submission. These are AnimaGuru credits; compare returned amounts directly with the authorized credit budget. Actual usage can differ from estimates.
+
+Use only a model and options returned by the live catalog. An image can establish an art direction or become a video reference. An illustrative prompt shell is:
+
+> Original tactile editorial paper collage: matte warm cream cardstock, cobalt blue and fox-orange cutouts, visible paper fibers, slightly imperfect scissor edges, soft short shadows falling down-right. [One concrete composition and action.] Restrained studio lighting and readable silhouettes. Keep a quiet area for separately composed editorial type. No lettering, captions, logos or watermark.
+
+Describe the subject, position, scale and material of each important element. Keep recurring character features verbatim between requests. Exact logos and text belong in the composition, not in model-generated pixels.
+
+## Submit and track
+
+For example, after replacing placeholders with discovered values and validating the settings:
+
+```sh
+animaguru assets generate PROJECT_ID --type image --name "cutcraft-keyframe" \
+  --model IMAGE_MODEL --aspect-ratio 3:2 --prompt "SAVED ART DIRECTION" --dry-run
+animaguru assets generate PROJECT_ID --type video --name "cutcraft-scene" \
+  --model VIDEO_MODEL --workflow image-to-video --reference IMAGE_FILE_SHA \
+  --seconds 8 --resolution 768p --mode speed \
+  --prompt "SAVED VISUAL ACTION" --dry-run
+```
+
+The literal values above are examples, not assumed capabilities. Image aspect ratios, video modes, duration limits, soundtrack handling and audio support vary by model. Add `--yes` only for the user's authorized work. Save the queued receipt immediately. Submission results may be job receipts rather than completed assets.
+
+Use `jobs list PROJECT_ID` and `jobs wait PROJECT_ID JOB_ID --timeout SECONDS` to track the original job. A timeout does not cancel it. Check the job and saved assets before resubmitting after an uncertain response. Do not repeatedly regenerate failed work without diagnosing the error and checking the remaining budget.
+
+When a job explicitly fails because its provider is unavailable, preserve that receipt and inspect models from another provider. A replacement with compatible capabilities and an estimate inside the authorized scope can continue the task. Remove unsupported options when changing providers; a seed or prompt-expansion setting valid for one model may be invalid for another. Relay the returned charge or refund status without assuming all failed jobs are free.
+
+Download finished assets with `assets download PROJECT_ID ASSET_ID --output FILE`. Keep stdout receipts separate from media and never overwrite an existing downloaded file inadvertently.
+
+## Choose continuity deliberately
+
+- For different explanatory ideas, use intentional cuts, a shared palette and a common layout. An unrelated next scene does not need the previous clip's last frame.
+- For one continuous action, use a supported video extension workflow with the saved video file SHA as `--source`. Inspect the catalog for its resolution and duration limits. `extendOutput` distinguishes the combined video from continuation-only output; inspect and trim any returned overlap rather than counting it twice.
+- First/last-frame workflows use `--reference` and `--end-reference` only when the model profile supports them. Both must be saved files belonging to the selected project.
+- A locally extracted end frame is not automatically a project asset. Use an explicitly supported upload/import path before referencing it; do not invent an extract-frame or raw upload MCP tool. A character-import command can redraw and charge credits, so it is not a transparent generic frame upload.
+
+Do not carry over the reference workflow's fixed ten-second clip assumption. Choose durations from the model profile and the story, and measure the returned files.
+
+## Narration and assembly
+
+Generate one narrator with `assets generate ... --type speech --voice VOICE --delivery INSTRUCTIONS --prompt TRANSCRIPT`, after discovering the available speech model and voice. Use delivery instructions supported by that provider. Keep the spoken transcript independent of the visual directions. Download and measure the actual recording with `ffprobe`.
+
+For generated motion layers, mute their sound when it conflicts with the separate narration. Reuse or generate a compatible instrumental bed at low level. Keep clip selection, crop, text, logos and all timing deterministic in the composition. Use an existing AnimaGuru HTML/Pixi film workspace when one is supplied; otherwise choose a local composition/export tool suited to the requested deliverable. Keep source files and a reproducible export command.
+
+For controlled cutout animation, request `transparent: true` image assets (CLI `--transparent`) and check the downloaded alpha channel. Keep a subject's complete silhouette inside the image. Animate these assets as separate layers against a fixed setting; drive position, rotation, entrances and holds from absolute time. Use actual transparent silhouettes for a moving subject instead of moving an opaque rectangular illustration. If video providers are all unavailable, this route can complete the film using working image and speech providers without repeated failed video submissions.
+
+Normalize frame size, frame rate and pixel format before concatenation. Export a broadly playable H.264/AAC MP4, or the user's requested format. Check duration, dimensions, audio presence, frame count and representative rendered frames. Preserve truthful asset attribution: motion made with AnimaGuru remains generated footage even when exact typography and editing are composed locally.

package/skills/animaguru-video/SKILL.md

@@ -21,6 +21,8 @@ Read the sibling `animaguru` skill for authentication, model discovery, credit a
 
 Keep the gallery URL in `galleryUrl`. All available style skills are already installed for both agents; selection does not require another CLI command, download, account or paid generation.
 
+An explicit request for the installed `animaguru-art-motion` skill is a custom style choice. Read its `SKILL.md`, select the relevant art material and/or explainer reference, and continue the shared brief without asking for another gallery choice. Save `skill` as `animaguru-art-motion` and leave `template` null; record the selected reference names in `production/brief.md`. The skill contains 36 art treatments and nine motion grammars as references, rather than additional gallery presets.
+
 
 ## Purpose workflows and saved projects