AI BriefingAnatomy: Explaining Concepts with Moving Machines Instead of Boxes and Arrows
·2026.11.05 15:30
Key point
Anatomy visualizes concepts as moving machines for use in technical documentation and lecture materials.
Details
- Released an agent skill called 'Anatomy' that generates diagrams of hypothetical machines moving according to the concept, instead of box-and-arrow diagrams.
- Explains complex concepts such as LRU caches, Bloom filters, and GPU contexts by converting them into intuitive machines like bookshelves, slot machines, and gantry plotters.
- Outputs in the form of Node.js, HTML, SVG, and React components, supporting pointer and keyboard interactions as well as real-time numerical displays.
- Written for Claude Code, but complies with the Agent Skills format, allowing it to run on Antigravity, OpenCode, Codex, and MiniMax Code.
- Complex structures may require over 10 hours of agent work time, such as 1 hour 54 minutes for a clock movement and 10 hours 33 minutes for a Raptor engine.
- Automatically verifies overlap and pipe penetration issues using geometry audit and line-end inspection tools, reducing pre-deployment errors to 0.
- Distributed as the npm package 'anatomy-figures' under the MIT license.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Agent Standup: Open-source tool visualizes Claude Code sub-agents as a live team meeting
Claude Code skill converts arXiv papers to translated books preserving LaTeX structureOpenQodex: Open-Source AI Code Review Tool Verifies Agent Changes with 98% Issue Recall