Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Defuddle Leitura Limpa

ASecurity

Sanitize a web page to relevant content only before giving it to the agent (remove chrome/navigation/ads/scripts) to save context. Use when the user says "limpar página", "sanitizar HTML", "só o conteúdo", "página muito grande", "estourou contexto com web", or in EN "sanitize page", "clean HTML", "remove boilerplate", "defuddle page", "strip chrome/ads", "reduce page context".

2 stars
0 votes
0 copies
0 views
Added 9/19/2026
ai-agentsgoapi

Works with

api

Security Analysis

A100/100

Scanned 9/19/2026

Install to Claude Code

$npx -y skills add majinmagros/magros.ai-skills --skill defuddle-leitura-limpa --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Defuddle Leitura Limpa?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Defuddle Leitura Limpa
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majinmagros-defuddle-leitura-limpa/badge)](https://www.skillsdirectory.com/skills/majinmagros-defuddle-leitura-limpa)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
name: defuddle-leitura-limpa
description: Sanitize a web page to relevant content only before giving it to the agent (remove chrome/navigation/ads/scripts) to save context. Use when the user says "limpar página", "sanitizar HTML", "só o conteúdo", "página muito grande", "estourou contexto com web", or in EN "sanitize page", "clean HTML", "remove boilerplate", "defuddle page", "strip chrome/ads", "reduce page context".
---

# Defuddle — Leitura Limpa

Antes de entregar uma página web ao agente, reduza-a ao conteúdo
relevante: remove chrome, navegação, anúncios e scripts, preservando o
principal + metadados. Conceito da skill Defuddle (Maestros da IA, vídeo V1nMBWzXsCI).

## Quando usar / When to use

- "Pega o conteúdo dessa URL" e a página é grande / cheia de menus e ads.
- O agente estourou contexto lendo HTML cru.
- Pipeline de scraping/RAG antes de indexar ou resumir.
- Qualquer fetch web cujo HTML bruto > ~30 KB de texto útil estimado.

## Passos

1. **Extrair o bruto** — obtenha HTML ou texto da página (fetch, headless ou
   texto já coletado). Guarde URL original, título da aba/documento e data
   de coleta. Não resuma ainda.
2. **Remover boilerplate** — descarte, nesta ordem:
   - `<script>`, `<style>`, `<noscript>`, iframes de ads/trackers, SVGs decorativos;
   - `header`, `footer`, `nav`, sidebar, menus, breadcrumbs, paginação;
   - banners de cookie/LGPD, popups, paywall teasers, "assine", social share;
   - comentários, related-posts, "leia também", footers de SEO.
   - Regra: na dúvida entre menu e artigo, mantenha o artigo.
3. **Preservar conteúdo principal + metadados** — o output deve conter:
   - `título`, `URL`, `data de publicação` (se houver), `autor` (se houver);
   - corpo principal em ordem de leitura (h1–h3, parágrafos, listas, tabelas);
   - alt-text de imagens informativas; links relevantes como `[texto](url)`.
   - Se a página tiver múltiplos artigos (home, categoria), extraia só o
     solicitado ou liste títulos + links — nunca concatene tudo.
4. **Medir a redução** — reporte sempre:
   - `chars antes → chars depois (% redução)`;
   - estimativa de tokens (~chars/4) antes/depois;
   - o que foi removido (ex.: "nav+footer+3 ads+2 scripts").
   - Se redução < 20%, declare: página já era limpa ou extrator falhou.

## Regras

- NUNCA invente conteúdo para preencher buracos do boilerplate removido.
- NUNCA descarte tabelas, números, datas ou citações do corpo principal.
- NUNCA envie HTML cru ao modelo se a versão limpa já existe.
- Se o conteúdo principal não for identificável (JS-only, CAPTCHA, login),
  declare o bloqueio em vez de retornar chrome residual.
- Uma página = um documento limpo. Não misture páginas sem separar por URL.

## Output format

```text
Fonte: <título> — <URL> (<data>)
Redução: <antes> → <depois> chars (~<tokens antes> → ~<tokens depois> tokens, -X%)
Removido: [nav, footer, ads, scripts, ...]

<conteúdo principal em markdown>
```

## Related skills

- `exa-search` — buscar antes de limpar; limpe só o que vale ler.
- `data-scraper-agent` — pipeline de coleta em escala; defuddle é a etapa de limpeza.
- `context-budget` — quando a redução não basta e é preciso orçar contexto.

Attribution

majinmagrosmajinmagros
View sourceMore from majinmagros →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.

1023331 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

686011 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3331 votes

catchup

Recovers prior coding-agent session context by running `catchup <agent> --since-compact`, which extracts a clean summary of a previous Codex, Claude Code, Antigravity, OpenCode, or Pi Agent session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", or asks to recover/summarize a previous session before continuing. Do NOT use for the current conversation, git history, or any non-agent log.

611 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →