AI Inspiration WorldAdded Nov 3, 2026
cell-image-skill
Turns papers, abstracts and research plans into mechanism diagrams and roadmaps, delivered as editable SVG plus PNG.
- Ecosystem
- Agent Skills
- Skill category
- Research
- Creator / organization
- butterflylittle（GitHub 用户）
What it is
cell-image-skill is a scientific figure Skill by GitHub user butterflylittle; the README calls it scientific visualization infrastructure for AI agents. Give it a paper, abstract, project or research plan, and it first works out from the evidence what the figure should show, then draws a mechanism diagram or technical roadmap. By default you get an editable SVG and a PNG, run through its own structure and SVG checkers. The Skill forbids generating or faking real experimental images, pathology slides or medical scans. It makes research schematics and doesn't diagnose anything.
Who it's for:Researchers who need mechanism diagrams and roadmaps for papers and talks.
Before you start: Clone it with Git into ~/.agents/skills/cell-image and call it in Codex with $cell-image; the README only gives install steps for Codex. The checkers suggest Python 3.9 or later, and the full checks also need jsonschema, Pillow or Inkscape. The image-assisted route uses your AI's built-in image tools, and any cost depends on that service. The README is in English and the Skill itself is in Chinese. Check separately whether your journal accepts AI-made figures. Apache-2.0 license.
Based on the project's public documentation. Not tested by AI Inspiration World.