web-to-markdown

name	web-to-markdown
metadata	{"version":"1.1.0"}
description	Convert URLs or local files (PDF/DOCX/PPTX/HTML) to Markdown, with optional summarization. Single or batch processing. Uses markitdown for intelligent extraction, falls back to curl+pandoc. Use when fetching web pages as markdown, converting documents for analysis, or producing quick summaries.

Web to Markdown

Convert URLs and local files to Markdown for analysis, quoting, or processing. Optionally summarize long documents before deeper work.

When to use

Use this skill when you need to:

• pull down a web page as a document-like Markdown representation
• convert binary docs (PDF/DOCX/PPTX) into Markdown for analysis
• quickly produce a short summary of a long document before deeper work

Quick Usage

Convert a URL or file

uvx --from 'markitdown[pdf]' markitdown <url-or-path>

Save to a specific file:

uvx --from 'markitdown[pdf]' markitdown <url-or-path> > /tmp/doc.md

Convert + summarize

Use the wrapper script for summarization with context:

node scripts/to-markdown.mjs <url-or-path> --summary --prompt "Focus on security implications"
node scripts/to-markdown.mjs <url-or-path> --tmp     # write to temp file, print path
node scripts/to-markdown.mjs <url-or-path> --out notes.md

The summarizer:

1. Converts to Markdown via markitdown
2. Saves full Markdown to a temp file (prints path as hint so you can inspect the full content)
3. Runs pi --model claude-haiku-4-5 to summarize with your prompt

Always provide context — summaries are only useful when you provide what you want extracted and the audience/purpose. Without a focus prompt, summaries become generic and unhelpful.

Batch processing (shell function)

Add to .bashrc / .zshrc:

web-to-markdown() {
  local url output_dir format batch_file output_file
  output_dir="." format="markdown"

  while [[ $# -gt 0 ]]; do
    case $1 in
      --batch) batch_file="$2"; shift 2 ;;
      --output-dir) output_dir="$2"; shift 2 ;;
      --format) format="$2"; shift 2 ;;
      --output) output_file="$2"; shift 2 ;;
      *) url="$1"; shift ;;
    esac
  done

  _fetch() {
    local u=$1 fmt=$2
    local content
    content=$(uvx --from 'markitdown[pdf]' markitdown "$u" 2>/dev/null) \
      || content=$(curl -sL "$u" | pandoc -f html -t markdown)
    [[ -z "$content" ]] && return 1
    [[ "$fmt" == "plain" ]] && echo "$content" | pandoc -f markdown -t plain || echo "$content"
  }

   [[ -n  ]]; 
    [[ ! -f  ]] && {   >&2;  1; }
     -p 
     ok=0 fail=0
     IFS=  -r u || [[ -n  ]]; 
      [[ -z  ||  =~ ^# ]] && 
       name ext out
      name=$(  | sed  |  -c1-100)
      ext=$([[  ==  ]] &&  txt ||  md)
      out=
       _fetch   >  2>/dev/null; 
          >&2; (( ++ok ))
      
          >&2;  -f ; (( ++fail ))
      
     < 
      >&2
     $(( fail >  ?  :  ))
  

  [[ -z  ]] && {   >&2;  1; }
  [[ -z  ]] && {
     name ext
    name=$(  | sed  |  -c1-100)
    ext=$([[  ==  ]] &&  txt ||  md)
    output_file=
  }
  _fetch   >  &&   || {  -f ;  1; }
}

Examples:

web-to-markdown "https://example.com/article" --output article.md
web-to-markdown --batch urls.txt --output-dir ./docs --format markdown
web-to-markdown "https://example.com" --format plain

Dependencies

• markitdown (auto-installed via uvx) — smart content extraction
• curl + pandoc — fallback for when markitdown fails
• pi — only needed for --summary mode