doc-loader

Charge la documentation d'un framework depuis son site web dans des fichiers markdown locaux. Supporte Symfony, API Platform, Meilisearch, atournayre-framework et Claude Code.

⭐6stars🔀0forks📁View on GitHub🕐Updated Jan 2, 2026⬇️1downloads

When & Why to Use This Skill

The Documentation Loader skill automates the retrieval and conversion of web-based framework documentation into local Markdown files. Designed for developers, it supports popular ecosystems like Symfony and API Platform, providing a streamlined way to maintain up-to-date, searchable, and offline-accessible technical references directly within a local workspace.

Use Cases

  • Offline Development: Download complete framework documentation to maintain productivity in environments with limited or no internet connectivity.
  • AI Context Integration: Convert web docs into local Markdown files to serve as high-quality context for AI coding assistants, improving the accuracy of code generation and debugging.
  • Version Management: Fetch and organize documentation for specific framework versions to ensure technical references perfectly match the project's dependency requirements.
  • Automated Documentation Sync: Leverage the built-in 24-hour cache and scraping workflow to automatically keep local technical manuals synchronized with the latest official updates.
namedoc-loader
description>
allowed-tools[Task, WebFetch, Write, Edit, Bash, Read, Glob]
modelclaude-sonnet-4-5-20250929

Documentation Loader Skill

Usage

/doc:framework-load <framework> [version]

Frameworks supportés

Framework Agent Path
symfony symfony-docs-scraper docs/symfony/[version]/
api-platform api-platform-docs-scraper docs/api-platform/[version]/
meilisearch meilisearch-docs-scraper docs/meilisearch/[version]/
atournayre-framework atournayre-framework-docs-scraper docs/atournayre-framework/[version]/
claude claude-docs-scraper docs/claude/

Workflow

  1. Parser arguments (framework, version)
  2. Vérifier README contenant les URLs
  3. Gérer cache (24h par défaut)
  4. Déléguer à agent scraper spécialisé
  5. Générer statistiques finales

Cache

  • Fichiers < 24h : ignorés
  • Fichiers > 24h : supprimés et rechargés
  • Délai 2s entre requêtes (anti rate-limit)

Error Handling

  • Framework non supporté → ARRÊT
  • README introuvable → ARRÊT
  • Échec scraping URL → WARNING (continue)

References

Notes

  • Fichiers markdown nommés depuis URL
  • Support multi-version