# cathrynlavery/diagram-design

Source: [GitHub](https://github.com/cathrynlavery/diagram-design)  
Feed7 permalink: https://feed7.dev/p/diagram-design-1456lje  
Published: Unknown  
Trust: Needs Review (needs_review)

## Why Included

Diagram Design gives coding agents a reusable system for branded, accessible HTML and SVG diagrams, including draw.io and Mermaid redraw workflows.

## Source Summary

The repository packages **27 diagram types** in **three visual variants** as self-contained HTML and SVG. It can derive brand tokens from a website and redraw draw.io or Mermaid sources for different formats and detail levels.

## Practical Implication

Agent users can replace one-off diagram prompting with a skill that separates type selection, brand tokens, import fidelity, accessibility, and export. Its progressive references also keep irrelevant diagram instructions out of context.

## Agent-Ready Context

The repository packages **27 diagram types** in **three visual variants** as self-contained HTML and SVG. It can derive brand tokens from a website and redraw draw.io or Mermaid sources for different formats and detail levels.

Agent users can replace one-off diagram prompting with a skill that separates type selection, brand tokens, import fidelity, accessibility, and export. Its progressive references also keep irrelevant diagram instructions out of context.

The fetched page is internally inconsistent: its header says **29 types**, while the body repeatedly documents 27. PNG export also requires Playwright, and customized packaged files may be overwritten by updates.

## Connected Context

Feed7 judgment across 461 accumulated Signals:

This specializes the portable-skill pattern for diagram production, replacing unconstrained prompting with explicit choices for diagram type, branding, source import, accessibility, detail, and export. Its progressive references reinforce selective context loading, while the count inconsistency, Playwright dependency, and update-overwrite risk show why skill packages need verified documentation, dependency checks, and separation of user customization from managed assets.

- [Skills are new features: Building Skill-Centric Harness — Yogendra Miraje, FactSet](https://feed7.dev/p/skills-are-new-features-building-skill-centric-harness-yogendra-miraje-f-0lp4c7o) — The repository exemplifies progressive skill loading, while its dependencies, conflicting documentation, and update behavior support treating skills as versioned products requiring ownership, evaluation, and lifecycle controls.
- [Leonxlnx/taste-skill](https://feed7.dev/p/taste-skill-15nf4kv) — Both encode visual judgment as portable agent instructions, but diagram-design narrows that approach into structured diagram selection, brand-token extraction, import fidelity, accessibility, and export.
- [Nutlope/hallmark](https://feed7.dev/p/hallmark-0f5n823) — Hallmark applies selection and critique to generated interfaces; diagram-design applies a comparable reusable quality layer to diagrams and adds source-format transformation.
- [browser-use/video-use](https://feed7.dev/p/video-use-1sjph9l) — Both replace broad media prompting with staged, bounded workflows and explicit rendering dependencies, showing how specialized skills can separate planning, inspection, and export.

## Context Map

- Layer: agent
- Domains: coding, image
- Topics: skills, context-engineering, design-engineering

## Uncertainty

- The fetched page is internally inconsistent: its header says **29 types**, while the body repeatedly documents 27. PNG export also requires Playwright, and customized packaged files may be overwritten by updates.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.