od-contribute
One-click contribution flow for OpenDesign (nexu-io/open-design) — even for non-coders. Pick…
Apply motion craft to an active task: refine a brief, beat map, visual continuity, typography, pacing, and timeline implementation. Use when planning or improving motion graphics; the host task owns artifact creation, output format, and export.
$ npx -y skills add nexu-io/open-design --skill motion-design --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/motion-designContext preview
The summary Claude sees to decide when to auto-load this skill.
Apply motion craft to an active task: refine a brief, beat map, visual continuity, typography, pacing, and timeline implementation. Use when planning or improving motion graphics; the host task owns artifact creation, output format, and export.
name: motion-design description: "Apply motion craft to an active task: refine a brief, beat map, visual continuity, typography, pacing, and timeline implementation. Use when planning or improving motion graphics; the host task owns artifact creation, output format, and export." triggers: - motion design - motion graphics - 视频动效 - 动态设计 - MG动画 - kinetic typography - animated explainer od: mode: utility category: animation-motion featured: -100 example_prompt: Create a 15-second motion film for a focus timer. Show scattered thoughts converging into one clear task, with expressive typography, a continuous visual motif, and a confident closing frame.
Use this functional helper to refine the user's brief or work in progress with one clear idea expressed through composition, transformation, rhythm, and appropriate sound unless the user requires silence. Apply the relevant craft guidance within the current task; this skill is not a rendering template and does not create a project or choose an output format. The Motion Design creation route owns those decisions. Quality means viewers understand and remember the idea while wanting to watch it move.
Follow the active host's task route, stage, capability, and delivery contract. This skill supplies motion craft; it does not create tool access or change the host's workflow. In OD Next, retain its Core, general orchestration, keyed production-ready marker, and ship-on-write default. Plan in ordinary language; author during production. Never emit Plan Contract or Runtime State blocks, request protocol repair, or claim a host-frozen delivery contract. Only actual user-requested video-motion production, when the host Core authorizes it, permits the bounded video check below. For a video brief (a timed film, promo, explainer or showreel), declare BOTH editable HTML source and MP4 as required deliverables. Rendering the MP4 is Build production and must finish before delivery. For pointer/scroll-driven interaction or UI micro-animation, deliver runnable HTML/code instead; do not flatten the requested interaction into video. Never describe an unrendered video file as complete.
Where the active task's tools and scope permit visual iteration, examine the opening, key transformation, transition, text hold and ending at intended size, plus continuous playback. Correct the largest observable defect, then revisit only affected moments. A runnable file alone is not proof of successful motion. Distinguish what was observed from what was only specified; never invent quality measurements.
Use the user's assets, references, script and previous decisions first. Resolve: message and audience; destination/aspect; duration; exact copy; must-show assets; visual direction; audio; and required output. Ask only when missing information would change the story or make the result misleading. Otherwise choose a coherent direction and state the meaningful assumption briefly. Do not impose three style options or an approval round on a user who asked you to make the piece.
For an unconstrained short film, a useful starting point is 12–18 seconds, 16:9, 1920×1080, 30fps, no invented voiceover, and a replayable source. A portrait social brief calls for 9:16. Existing brand, timing, platform and copy override these defaults. Do not invent product claims, metrics, testimonials or assets.
Create a compact beat map as part of the existing plan, not a second document: `time → viewer takeaway → focal object → transformation → copy/audio → exit`. Time budget includes entrances, reading holds and exits. Choose a recurring object, shape, line or typographic motif that connects the beats.
Example mechanism for a focus product (adapt, do not copy as a universal style): scattered words orbit → one word attracts the others → orbit becomes a timer → timer completes and becomes the wordmark. Each movement changes the meaning; four unrelated title cards would lose the mechanism.
Resolve one strong representative frame in the composition before expanding all scenes. Choose the actual hierarchy, crop, type, palette, texture and focal object. A weak still will not be rescued by particles, camera shake or effects. When planning-only, describe this frame; do not create an artifact early.
and "modern" alone are not art directions. Define what fills the frame, where attention rests, the material language and the contrast structure.
negative space and font character. Choose fonts that cover the actual language; use tested fallbacks and wait for font readiness before capture.
glass cards, glowing orb or beige background independent of the brief.
paper, tactile collage, vector geometry, painterly field, or another justified system. Carry its lighting, edges and textures through the film.
a headline above three cards. Keep subtitles and platform chrome out of the focal area; use destination-specific safe margins, not one universal ratio.
Crop deliberately rather than stretching. Keep data axes, units and values correct through their entire animation.
Choose the lead action and let supporting motion answer it. Offset independent objects when that clarifies reading; move them together when collective motion is the idea. Do not mechanically stagger every word or enforce a fixed moving fraction of the screen.
Use three possible layers, not three mandatory eff
The 1st Collaborative AI Design Agent Workspace. Local-first and open source. Create prototypes, websites, apps, dashboards, presentations, images, and videos with your team and AI agents. Share work, exchange feedback, and iterate together in one workspace.
Repo: nexu-io/open-design
One-click contribution flow for OpenDesign (nexu-io/open-design) — even for non-coders. Pick…
Hyperframes-based video template for retro pixel deck motion design. Use when users want a…
Generate and iterate ad creative including headlines, descriptions, and primary text. Useful…
Luxury dark-editorial HyperFrames template for three-page cinematic storyboards, inspired by…
Browser automation CLI for AI agents. Use when the user needs to inspect, test, or automate…
Full-lifecycle AI music album production — concept, lyric drafting, track sequencing, and…