convez.ai
SkillProdutividadeIntermediário
chrisjianghp

md2mindmap

chrisjianghp/md2mindmap

Converte Markdown em mindmaps interativos com zoom, pan e exportação PDF

Ver no GitHub
InicialPluginScore · de 0 a 100

Uma skill Claude Code que transforma documentos Markdown (locais ou online) em mindmaps interativos em HTML (com exportação PDF) ou documentos Lark/Feishu. Ideal para visualizar estruturas de arquivos, documentações e designs de skills de forma rápida e exploratória.

Instalação

$npx convez-ai chrisjianghp/md2mindmap

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
2
Forks
1
Acompanhando
0
Saúde
42%
Último commit
há 5 meses
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Visualizar estrutura de documentações e README.md em mindmaps interativos
  • Exportar mapas mentais para PDF direto do navegador
  • Integrar mindmaps em documentos Lark/Feishu para equipes
  • Analisar arquitetura e design de skills do Claude Code

Veredito

Skill bem implementada que resolve um problema real: transformar markdown em visualizações mentais. Suporta múltiplos formatos e plataformas (HTML, Lark/Feishu), mas requer setup adicional para funcionalidades Lark. Ótima para quem trabalha com documentação técnica e precisa de visualizações rápidas.

README

A tradução para o português deste README ainda está na fila.

md2mindmap

A Claude Code Skill that converts a Markdown document — local file or online URL — into an interactive mindmap.

It can output:

  • HTML Markmap — a local interactive HTML file with zoom/pan/fold and PDF export
  • Lark/Feishu document — a freshly-created Lark doc containing a Mermaid mindmap whiteboard
  • Both

What it does

Two input modes:

  • Mode A — Local MD file: Path like ~/path/to/README.md or ./docs/intro.md. Read with the Read tool.
  • Mode B — Online MD URL: A URL pointing to raw markdown — GitHub raw links (https://raw.githubusercontent.com/...), GitHub blob pages (auto-converted to raw), or any HTTP(S) link returning markdown text. Fetched with WebFetch.

Before generating, the skill resolves three preferences:

  1. Language: keep the source language or translate to Chinese
  2. Depth: title skeleton only, or titles + short summaries
  3. Format: HTML file, Lark/Feishu document, or both

If the user already specified any of these in the prompt (for example "生成 HTML", "仅标题", "保持原文"), the skill respects that and only asks about missing choices.

HTML output

The HTML output uses Markmap via markmap-autoloader.

Features:

  • Interactive zoom, pan, fold/unfold
  • Dark theme with bright readable text
  • Default first-level expansion (initialExpandLevel: 1) for a clean overview
  • PDF export button
    • Uses html2canvas to snapshot the rendered mindmap into a PNG
    • Opens the browser print dialog; choose Save as PDF
    • This avoids Chrome's SVG foreignObject / Chinese font issues when printing

Lark/Feishu output

The Lark/Feishu output creates a new Lark document, inserts a blank whiteboard, then writes a Mermaid mindmap into that whiteboard.

This output requires lark-cli authentication.

Not in scope

This skill handles plain Markdown text only. It does not parse Lark / Notion / Confluence / other rich-text systems directly.

Installation

Clone into your Claude Code skills directory:

git clone https://github.com/chrisjianghp/md2mindmap.git ~/.claude/skills/md2mindmap

That's it. Claude Code will pick up the skill on the next session.

Requirements

  • Claude Code with skills support
  • For HTML output: internet access to load CDN resources (markmap-autoloader, html2canvas)
  • For Lark/Feishu output:
    • lark-cli — authenticated (lark-cli config init)
    • Companion skills (usually bundled together): lark-shared, lark-doc, lark-whiteboard

Usage

Just ask Claude things like:

The skill triggers automatically based on your phrasing.

Repository layout

md2mindmap/
├── SKILL.md            # The skill definition (what Claude reads)
├── .gitignore
├── LICENSE
└── README.md

License

MIT

Ver o original no GitHub

Bom para

Relacionados