convez.ai
SkillProdutividadeIntermediário
Jyo238

Report Kit

Jyo238/report-kit

Relatórios HTML portáteis, editáveis e sem dependências

Ver no GitHub
InicialPluginScore · de 0 a 100

Skill para Claude Code que gera relatórios HTML autossuficientes e editáveis no navegador. Destinatários podem visualizar, editar, salvar e exportar para PNG sem instalar nada — apenas abrem o arquivo .html no Chrome ou Edge.

Instalação

$npx convez-ai Jyo238/report-kit

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

Casos de uso

  • Gerar relatórios técnicos compartilháveis via email ou Teams
  • Criar dashboards e cards interativos que qualquer pessoa pode editar
  • Exportar análises para PNG mantendo formatação exata

Veredito

Solução prática para distribuição de relatórios sem fricção. Combina bem com fluxos onde a edição colaborativa e portabilidade são críticas, embora a fidelidade de exportação PNG chegue a ~95%.

README

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

report-kit

Generate self-contained HTML reports that anyone can open in a browser to view, edit, save, and export to PNG — with nothing to install.

It's a Claude Code skill: the report author uses an AI agent to produce the report; the finished .html is then shareable to anyone with a browser.

Two audiences, two requirements

WhoWhat they doWhat they install
Author (writes reports)Has Claude Code + this skill; asks the agent to make a reportthe skill + Python 3 (for the bundler)
Recipient (reads/edits)Opens the .html in Edge/Chromenothing

Install (author side)

Copy the report-kit/ folder into your personal skills directory:

  • Windows: C:\Users\<you>\.claude\skills\report-kit\
  • macOS/Linux: ~/.claude/skills/report-kit/

The folder is self-contained — it already bundles assets/html2canvas.min.js, so no internet is needed at build time. Then just ask your agent, e.g. "用 report-kit 幫我做一份 X 的報告".

What the finished report can do

Open it in Edge or Chrome (ideally served from localhost). Bottom-right:

  • ✏️ Edit — click into the page and change text.
  • 💾 Save (Ctrl+S) — write changes back to the same .html (File System Access API).
  • 📷 Save as Image — export the page to PNG, pick where to save.

All browser-native + bundled; works offline; no extensions, no tools.

How it works (no dependencies for recipients)

  • Edit/Save use the browser's File System Access API (Edge/Chrome).
  • Save-as-Image uses html2canvas, embedded inline in every report (no CDN).
  • Result: one portable file. Email it, drop it in Teams/SharePoint — it just works.

Notes

  • Save-as-Image fidelity is ~95% (great for cards/tables/badges). For pixel-perfect exports, an author can use the companion html-to-image skill (Playwright) — not needed by recipients.
  • file:// double-click works, but for reliable Save use a local server (e.g. VS Code Live Server) so the page runs on http://localhost.
  • Firefox lacks the File System Access API → Save/Image fall back to downloads.

License

MIT — see LICENSE. Bundles html2canvas (MIT) for in-browser PNG export.

Ver o original no GitHub

Bom para

Relacionados