read-webpage-content-as-markdown

Read a webpage into cleaned markdown using curl + markitdown + codex exec. Use whenever asked to read a webpage or extract article content from a URL. Static HTML only; JS/client-rendered pages require a Playwright workflow.

⭐0stars🔀0forks📁View on GitHub🕐Updated Jan 5, 2026

When & Why to Use This Skill

This Claude skill enables the efficient conversion of static webpages into clean, structured Markdown format. By utilizing curl and markitdown, it extracts core article content while removing unnecessary HTML clutter, making it an essential tool for web-based data gathering, content analysis, and providing high-quality context to AI models.

Use Cases

  • Article Extraction: Quickly transform online news, blog posts, and reports into distraction-free Markdown for summarization and analysis.
  • Documentation Fetching: Retrieve technical documentation from static websites to serve as clean, formatted context for programming and technical tasks.
  • Knowledge Base Population: Streamline the process of gathering web information into a structured Markdown repository for research and archival purposes.
  • Content Cleaning: Automatically strip away site-wide navigation, ads, and scripts to focus exclusively on the primary text and relevant in-page links.
nameread-webpage-content-as-markdown
descriptionRead a webpage into cleaned markdown using curl + markitdown + codex exec. Use whenever asked to read a webpage or extract article content from a URL. Static HTML only; JS/client-rendered pages require a Playwright workflow.

Read Webpage Content as Markdown

Use:

scripts/read-webpage-content-as-markdown.sh [--navlinks] <url> [output_md]

Notes:

  • Uses curl (static HTML only); JavaScript is not executed.
  • Temp artifacts are stored under /tmp.
  • Output includes YAML frontmatter: source_url, accessed_at, commands.
  • Output path defaults to /tmp/read-webpage-content-as-markdown.<timestamp>.md; relative output paths are written under /tmp/.
  • --navlinks keeps only topic-relevant navigation links (e.g., in-page table of contents); it drops site-wide menus and unrelated links.
  • If the script reports JS/client rendering, retry with Playwright.