Skip to content

Latest commit

 

History

History

README.md

🧪 Examples Index

This folder contains self-documenting examples demonstrating Semantic JSON's compilation, import, and export features.

📁 File Index

File Type Purpose
conformance-test-card.canvas JSON Canvas Primary test file. Nodes literally describe their expected positions—the file documents itself.
conformance-test-card.json Pure JSON Exported version with spatial metadata stripped. Compare to .canvas to see what gets removed.
conformance-test-card.png Screenshot Visual reference showing how the test card renders in Obsidian.
before-after.png Screenshot Hero image comparing scrambled vs semantic JSON ordering.
🩷Catppuccin.pure.json Pure JSON Real-world example: Catppuccin color palettes exported for LLM consumption.

🔬 How to Use These Examples

Test the Conformance Card

  1. Copy conformance-test-card.canvas into your Obsidian vault
  2. Open it in Obsidian Canvas
  3. Make any change and save
  4. Observe: Obsidian scrambles the node order
  5. Install Semantic JSON plugin
  6. Save again
  7. Observe: Order is restored to semantic sequence

Understand the Transformation

Compare these two files side-by-side:

conformance-test-card.canvas  →  conformance-test-card.json
         ↓                                ↓
   Full Canvas format              Pure data artifact
   (x, y, width, height)          (only semantic content)
   (color, fromSide, toSide)      (edges simplified)

The .json is what you'd feed to an LLM or use in a pipeline.

What the Test Card Tests

  • Spatial ordering: Top-left to bottom-right reading sequence
  • Hierarchical grouping: Groups followed by their children
  • Nested groups: Child groups within parent groups
  • Color taxonomy: Edges colored by direction (horizontal vs vertical)
  • Mixed node types: Text nodes, link nodes, groups
  • Edge topology: Arrows ordered by connected node positions

🩷 About the Catppuccin Example

🩷Catppuccin.pure.json demonstrates:

  • Unicode filename handling (the pink heart emoji)
  • Real-world content (color palette definitions)
  • LLM-ready format (clean JSON, no rendering metadata)
  • Multiple palette variants (mocha-frappe, rose, sage, grape, honey, etc.)

This is the kind of output you'd share with an AI assistant or consume in a data pipeline.

🔗 Related Methodologies

This self-documenting examples pattern aligns with broader documentation architectures:

  • Semantic JSON - Parent project: Obsidian plugin for smart Canvas data recompiling
  • Context Engineering Skills - Dual-format documentation system using similar conformance testing patterns
  • ZK Context Vault - Comprehensive agent system documentation following self-documenting principles

Back to main README