iart-ai/motion-skills

50 open-source skills that teach your AI coding agent to make motion graphics, animation & video — kinetic typography, data-viz, explainers, TikTok/Reels, WebGL, Manim. 14 installable packs. By iart.ai, the AI motion agent.

What it solves

This project provides a library of 53 open-source "skills" that enable AI coding agents to create professional-grade motion graphics, animations, and videos. It addresses the difficulty of getting timing, easing, safe areas, and frame-accurate rendering correct via simple prompts, which typically lack the craft of a professional motion designer.

How it works

Instead of fine-tuning models, the project uses a set of self-contained folders containing documentation (SKILL.md) and references. These skills teach agents how to follow professional workflows. For visual artifacts, it implements a "deliver-and-verify loop" where the agent renders a frame, takes a screenshot, and checks its own output using a provided toolkit of scripts to ensure accuracy.

Who it’s for

Content creators (TikTok, Reels, YouTube), marketers, frontend developers, and motion designers who use AI coding agents like Claude Code, Cursor, or Codex to automate the production of programmatic video and animation.

Highlights

  • Broad Motion Coverage: Supports everything from kinetic typography and data-driven charts to 3D/WebGL and Manim math animations.

  • Agent-Native Workflows: Designed to be installed as plugins or packs that are auto-discovered by 40+ AI agents.

  • Self-Verification: Includes scripts to freeze frames and probe MP4s so agents can inspect their own work.

  • Modular Installation: Organized into 16 specialized packs (e.g., tiktok-video-skills, web-animation-skills) so users only install what they need.

Related

  • Project
    Junchao-cs/SolarWM

    SolarWM is an open‑source framework that provides a massive, camera‑annotated video dataset, a three‑stage training recipe, and ready‑to‑use 5 B–33 B video world‑model checkpoints (Wan2.2, LTX‑2.5, MiniMax‑H3). It lets you train and run long‑horizon interactive video models with a unified CLI and fully documented pipelines.

  • Project
    taco-group/SparkVSR

    SparkVSR is an interactive video super-resolution framework that allows users to guide the restoration process using sparse high-quality keyframes to ensure better temporal consistency and controllable results.

  • Project
    op7418/guizang-product-video-skill

    The repository provides an AI‑driven “skill” that turns a software project’s recent updates into a short promotional video. It extracts changelog items, re‑uses the product’s UI components, writes copy, animates with GSAP/Three.js, synthesises music and sound effects, and finally exports an MP4 plus an editable video project. The skill is installed via the Vercel Skills CLI and works with Claude Code, Codex, or similar AI coding assistants.

  • Project
    simchowitzlabpublic/nano-world-model

    A minimalist framework for training video world models based on diffusion-forcing, used for future video prediction, 3D reconstruction, and robotic planning.