scientific-schematics

name	scientific-schematics
description	Advanced scientific diagram generator with AI-native iterative refinement. Use when creating Mermaid diagrams, scientific schematics, or technical visualizations.
license	Apache-2.0
metadata	{"version":"2.3.0","language":"en","author":"Antigravity"}

Scientific Schematics (Mermaid-Native)

Precision scientific visualization focusing on structural accuracy and universal accessibility.

🧠 Brain-Power: Structural Visualization

This skill prioritizes Structural Code (Mermaid) over raster images to ensure transparency, reproducibility, and high fidelity in technical environments.

1. Logic-First: We use the appropriate Mermaid diagram for the scientific concept.
2. Hybrid Fallback: generate_image is reserved for complex organic textures or artistic renderings that lack a structural schema.

🔄 The Master Workflow (Interactive)

Case A: Structural Blueprint (Primary)

When the concept can be modeled as a flow, hierarchy, or interaction:

1. Selection: Choose the optimal Mermaid type (see 📊 Mermaid Capability Map).
2. Generation: I generate valid Mermaid.js code with scientific themes.
3. Shadow Pipeline (Zero-Configuration Rendering):
   • Code saved to figures/[name].mmd.
   • Auto-Render: I run npx @mermaid-js/mermaid-cli to generate SVG (Journal) and PNG (Presentation).
4. Validation: I verify syntax and visual hierarchy.

Case B: High-Fidelity Artistic Proposals (Secondary)

When Mermaid cannot represent the complexity (e.g., a realistic anatomical heart):

1. Prompt Sharpening: Structured manifests for generate_image.
2. Generation: Create an image and store it in figures/[name].png.
3. Audit: Evaluation against Accuracy and Clarity.

📊 Mermaid Capability Map (Scientific Context)

Diagram Type	Best for...	Example
Flowchart (graph TD)	Decision trees, mechanisms, experimental protocols.	Enzyme-substrate interactions.
Sequence Diagram	Cascades, signal transduction, multi-step protocols.	Phototransduction pathway.
Class Diagram	Ontologies, taxonomies, data structures.	Phylogenetic trees or software architecture.
ER Diagram	Relational data, chemical bonds, networks.	Metabolic network relationships.
State Diagram	Cycle transitions, phase changes.	Cellular cycle or phase transitions.
Gantt / Timeline	Study durations, historical milestones.	Clinical trial stages.
Mindmap	Brainstorming, conceptual mapping.	Research field breakdown.
XY / Sankey	Trends, resource/energy flux.	Energy distribution in ecosystems.
Block Diagram	Modular architectures, lab setups.	Signal processing blocks.

🏢 Workspace Organization

Outputs should be placed in the project's target directory (e.g., docs/assets/ or assets/diagrams/) to maintain context.

[project-root]/
├── [target-dir]/            # Specified by task context
│   ├── [name].mmd          # Source Mermaid code (Editable)
│   ├── [name].svg          # Vector output (Publish Ready)
│   ├── [name].png          # Raster output (Static)
│   └── [name]_log.json     # Review & Metadata log

📊 Document Quality Standards

Type	Target	Logic
Journal	9.0+	SVG required. Zero ambiguity. Structural accuracy over aesthetics.
Technical	8.0+	High contrast, clear labels, academic palette (--theme neutral).

🌟 Guidelines for Antigravity

• Output Flex: Prefer absolute paths or paths relative to the project root for outputs.
• Theme Enforcement: Use --theme neutral as the default for scientific clarity.
• Node Detection: Check for node and npx before rendering. If missing, provide the .mmd code for the user's VS Code preview.