wheresryan22/anatomy

An agent skill for creating detailed, interactive isometric illustrations of real-world machines and complex systems. Generates polished SVGs with optional WebGL shaders and 3D effects.

What it solves

Anatomy is a tool for AI agents (specifically Claude, OpenCode, and Codex) that allows them to explain complex technical concepts by designing and building interactive, isometric figures. Instead of using generic diagrams, it enables the AI to invent a physical object whose mechanical working represents the technical idea (e.g., representing a cache key as a pin-tumbler lock).

How it works

The project provides a specialized drawing library (the "kit") and a set of craft rules that the AI agent follows. The workflow involves the AI collecting technical requirements, inventing a physical metaphor, planning the geometry, and using the the kit to generate the figure.

It supports three types of output:

  • SVG: Solid isometric parts with four-tone shading and eased motion.
  • SVG + WebGL: Combines SVG for structure and WebGL shaders for effects like fire, water, or glow.
  • 3D (Beta): Parts that can rotate and flex in true 3D space.

The system includes a verification suite that uses headless Chrome to audit geometry for overlaps, ensure lines don't end in mid-air, and verify the visual fidelity of the resulting figures.

Who it’s for

Developers and technical writers who use AI agents to create high-quality, interactive explanatory figures for documentation, blog posts, or landing pages.

Highlights

  • Metaphor-driven explanation: Converts abstract technical ideas into physical, working models.
  • Multi-modal rendering: Supports standalone HTML, SVG files, or React/Next.js components.
  • Automated verification: Includes a geometry audit and visual checks to ensure precision.
  • Zero-dependency kit: The core library uses only Node's built-in modules.
  • Agent-native: Designed as a standard Agent Skill for seamless integration with Claude Code, OpenCode, and Codex.

Written about in

Related

  • Project
    state-spaces/mamba

    Mamba is a linear-time sequence modeling architecture based on selective state space models (SSMs) that provides a high-performance, subquadratic alternative to Transformers.

  • Project
    lightgbm-org/LightGBM

    A highly efficient gradient boosting framework that uses tree-based learning algorithms to handle large-scale data with high speed and low memory usage.

  • Project
    tensorflow/tfjs

    A hardware-accelerated JavaScript library for training and deploying machine learning models in the browser and Node.js.

  • Project
    jlcodes99/cockpit-tools

    A universal account management tool for AI IDEs that enables one-click account switching, quota monitoring, and running multiple isolated instances of AI coding tools.