What problem does it solve?
Creating precise, editable, publication-quality architecture, workflow, and pipeline diagrams for academic papers is traditionally time-consuming, and AI-generated illustrative figures often lack determinism and editability, making them unsuitable for formal publications where consistency and fine-tuning are required.
Core Features & Use Cases
- Deterministic JSON-to-SVG Rendering: Identical FigureSpec JSON inputs always produce identical SVG outputs, ensuring consistency across revisions and collaborations.
- Shape-Aware Layout & Multilingual Support: Correctly clips edges to node boundaries for all supported shapes (rect, rounded, circle, ellipse, diamond) and accurately estimates widths for CJK multilingual labels.
- Editable Output for Academic Papers: Generated SVG files are plain-text and can be manually tweaked after generation, perfect for adjusting labels, colors, or layouts to meet journal or conference requirements.
- Use Case: For a MICCAI 2025 paper like FMC-Net, use this skill to generate the framework architecture diagram from a structured JSON spec, then edit node labels directly in the SVG if needed for clarity.
Quick Start
Use the figure-spec skill to generate an editable, publication-quality SVG diagram of your system architecture, workflow, or pipeline from a structured JSON specification of its components and connections.