name: fetch-url
short: Fetch and extract text content from URLs using lynx
long: |
  This command fetches and extracts plain text content from one or more URLs using the lynx text browser.
  It's particularly useful for automated content extraction, web scraping, and integration with LLM tools.
  
  INPUT: One or more URLs to fetch, with optional configuration for the output format and processing
  OUTPUT: Plain text content from the specified URLs, with configurable formatting
  
  The command uses lynx's dump mode to convert HTML content to plain text while:
  - Removing HTML formatting
  - Preserving text structure and hierarchy
  - Maintaining readable link references
  - Handling various character encodings
  
  Common use cases:
  1. Extract article content for summarization: --urls https://example.com/article
  2. Batch process multiple URLs: --urls https://site1.com,https://site2.com
  3. Custom formatting: --urls https://example.com --no-links
  
  Example outputs:
  ```
  Article Title
  ============
  
  Main content text here...
  
  References
  [1] http://reference1.com
  [2] http://reference2.com
  ```

flags:
  - name: urls
    type: stringList
    help: |
      URLs to fetch content from. Multiple URLs can be specified as comma-separated values.
      Each URL should be a valid HTTP/HTTPS URL.
      Example: --urls https://example.com,https://another.com
    required: true
  
  - name: no_links
    type: bool
    help: |
      If true, removes the reference links section from the output.
      Useful when you only want the main content without link references.
      Example: --no-links
    default: false

shell-script: |
  #!/bin/bash
  set -euo pipefail
  
  # Process each URL
  for url in {{ range .Args.urls }}"{{ . }}" {{ end }}; do
    echo "Fetching $url..."
    
    # Build lynx command
    lynx_cmd="lynx -dump "
    {{ if .Args.no_links }}
    lynx_cmd+="--nolist "
    {{ end }}
    
    # Execute fetch
    eval "$lynx_cmd '$url'"
    echo -e "\n---\n"
  done