skillZs
★ LIVE SKILL TAGS ★
>>> LIVE SKILLS INDEX <<<
* OPEN SOURCE *
NO LOGIN, NO TRACKING
※ REAL INSTALL DATA ※
← back to all skills
testacode/llm-toolkit55 installs

claude-md-writer

Escribe y mejora archivos CLAUDE.md siguiendo best practices de Anthropic. Este skill se activa cuando el usuario dice "crear CLAUDE.md", "mejorar CLAUDE.md", "actualizar CLAUDE.md", "revisar CLAUDE.md", "escribir instrucciones del proyecto", "create CLAUDE.md", "improve CLAUDE.md", "review CLAUDE.md", "write project instructions", "optimize docs for Claude", "auditar CLAUDE.md", "audit CLAUDE.md", "limpiar CLAUDE.md", "dead weight", o configura un nuevo repositorio.

How do I install this agent skill?

npx skills add https://github.com/testacode/llm-toolkit --skill claude-md-writer
view source ↗

Is this agent skill safe to install?

  • Gen Agent Trust Hubpass

    The claude-md-writer skill is a safe documentation utility designed to help users create and maintain CLAUDE.md files. It analyzes project structures and source files to provide relevant instructions, which constitutes a low-risk surface for indirect prompt injection. No malicious code, exfiltration, or dangerous command execution was found, and the skill explicitly promotes security best practices.

  • Socketpass

    No alerts

  • Snykpass

    Risk: LOW · No issues

  • Runlayerpass

    1/5 files flagged

What does this agent skill do?

CLAUDE.md Writer

A skill for creating and optimizing CLAUDE.md files following official Anthropic best practices.

Golden Rule

Para cada linea, pregunta: "Eliminar esto causaria que Claude cometa errores?" Si no, eliminalo. CLAUDE.md inflados causan que Claude ignore instrucciones.

"Think of CLAUDE.md as the 'unwritten knowledge' in your codebase"

Quick Tools

  • /init - Genera CLAUDE.md inicial basado en la estructura del proyecto
  • # key - Agregar instrucciones dinamicamente durante la sesion

@import Syntax

CLAUDE.md puede importar otros archivos:

See @README.md for overview and @package.json for commands.
Git workflow: @docs/git-instructions.md

Reference Documentation

For examples and validation checklist, see:

  • references/good-example.md - Well-structured CLAUDE.md example
  • references/bad-example.md - Common anti-patterns to avoid
  • references/checklist.md - Quality validation checklist
  • references/dead-weight-audit.md - Dead weight audit (5 criterios para recortar instrucciones)

What to Include vs Exclude

IncludeExclude
Bash commands Claude can't guessAnything Claude can figure out by reading code
Code style rules that differ from defaultsStandard language conventions
Testing instructions and preferred test runnersDetailed API docs (link instead)
Repository etiquette (branch naming, PR)Information that changes frequently
Architectural decisions specific to projectLong explanations or tutorials
Developer environment quirks (env vars)File-by-file descriptions
Common gotchas or non-obvious behaviorsSelf-evident practices

CLAUDE.md vs Skills

CLAUDE.mdSkills
Reglas amplias que aplican siempreConocimiento de dominio solo relevante a veces
Code style y workflowWorkflows especializados reutilizables
Convenciones del repoInfo que cambia por contexto
Se carga en cada sesionSe cargan on-demand

Content Triage

Antes de agregar contenido a CLAUDE.md, evaluar:

Si el contenido es...Entonces...
Regla que aplica a TODAS las tareasAgregar a CLAUDE.md
Domain knowledge especificoSugerir crear skill
Workflow que solo aplica a vecesSugerir crear skill
Info que cambia frecuentementeNo agregar, linkear a docs

Si el usuario quiere agregar info condicional ("Cuando trabajes con X...", "Para tareas de tipo Y..."), sugerir crear un skill separado.

Usage Modes

InvocacionComportamiento
/claude-md-writerProceso completo: analizar proyecto, escribir/mejorar CLAUDE.md
/claude-md-writer actualiza con la sesionExtraer info relevante de la sesion actual y agregar a CLAUDE.md
/claude-md-writer agrega regla: [descripcion]Agregar una regla especifica con el formato correcto
/claude-md-writer revisa y optimizaAuditar CLAUDE.md existente, eliminar redundancias, mejorar enfasis
/claude-md-writer auditaDead weight audit: aplicar los 5 criterios contra el setup actual (ver references/dead-weight-audit.md)
/claude-md-writer convierte a skillIdentificar contenido que deberia ser skill y extraerlo

Process

Step 1: Analyze the Project

Identify tech stack, project structure, and existing documentation.

Step 2: Apply the Three Dimensions

DimensionQuestion to Answer
WHATWhat is this project? Tech stack? Structure?
WHYWhy does it exist? What problem does it solve?
HOWHow do I work on it? Commands? Workflows?

Step 3: Write Following Best Practices

Length: Keep under 200 lines (ideal ~100 lines)

Structure (in order):

  1. Repository Overview (1 paragraph)
  2. Why This Exists (purpose, users)
  3. Quick Start & Commands (bash with descriptions)
  4. Architecture Overview (key patterns)
  5. Key Development Rules (with emphasis)
  6. References (links to detailed docs)

Emphasis for Critical Rules:

- **CRITICAL**: [Never violate this]
- **IMPORTANT**: [Always do this]

Progressive Disclosure:

  • Use file references instead of code snippets
  • Link to external docs for detailed topics
  • Keep CLAUDE.md focused on universal instructions

Step 4: Verify with Checklist

Before finalizing, run through the checklist in references/checklist.md.

Template

# CLAUDE.md

This file provides guidance to Claude Code when working with this repository.

## Repository Overview

[One paragraph: what this project is, tech stack, main purpose]

## Why This Exists

[What problem it solves, who uses it, business context]

## Quick Start & Commands

\`\`\`bash
npm ci        # Install dependencies
npm start     # Start dev server
npm test      # Run tests
npm run build # Production build
\`\`\`

## Architecture Overview

[Key patterns, folder structure, main concepts - keep brief]

## Key Development Rules

- **CRITICAL**: [Most important rule that must never be violated]
- **IMPORTANT**: [Second most important rule]

## References

- `/docs/` - Detailed documentation
- `README.md` - Project readme

Dead Weight Audit

Cuando el usuario pida auditar su setup, aplicar los 5 criterios de references/dead-weight-audit.md contra cada regla:

  1. ¿Default? — Claude ya lo hace sin que se lo digan
  2. ¿Conflicto? — Contradice otra regla en el setup
  3. ¿Redundancia? — Cubierto por otra regla o proceso
  4. ¿Parche puntual? — Agregado para arreglar un output malo específico
  5. ¿Vago? — Se interpreta distinto cada vez

Output esperado:

  • Lista de cortes con razón en una línea cada uno
  • Lista de conflictos encontrados entre archivos
  • Versión limpia del CLAUDE.md sin dead weight

Proceso post-audit: eliminar → probar 3 tareas típicas → re-agregar solo lo que se rompió.

Anti-Patterns to Avoid

Anti-PatternWhy It's BadDo This Instead
Code snippetsBecome stale, waste tokensUse file:line references
Styling rulesLLM is slow for deterministic tasksConfigure linters
Kitchen sinkDilutes instruction effectivenessKeep < 200 lines
No emphasisCritical rules same weight as minorUse CRITICAL/IMPORTANT
Auto-generatedMisses project nuancesManually craft
Domain knowledgeInflates every sessionMove to skills

Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.

<a href="https://skillzs.dev/skills/testacode/llm-toolkit/claude-md-writer">View claude-md-writer on skillZs</a>